SHEJEH SALAM (PROMOTING PEACE) BASELINE EVALUATION Baseline Evaluation | Shejeh Salam Activity 1 Activity Name: Shejeh Salam (Promoting Peace) Contract/Agreement Number: 72066820C00003 Activity Start and End Dates: September 30, 2020 to October I, 2025 Geographic Location: Juba, South Sudan COR Name & Office: Victor Lako / DG Submitted by: DT Global August 28, 2021 This publication was produced for review by the United States Agency for International Development. It was prepared by DT Global. DISCLAIMER The author’s views expressed in this publication do not necessarily reflect the views of the United States Agency for International Development or the United States Government. Baseline Evaluation | Shejeh Salam Activity ii CONTENTS Acronyms...................................................................................................................................... iii Table of Tables ..............................................................................................................................iv Summary Findings and Programmatic implications ................................................................. 1 Introduction......................................................................................................................................................... 1 Safety and Security ............................................................................................................................................. 1 Mechanisms and Capacity for Conflict Resolution, Mitigation, and Prevention.................................. 2 Social Cohesion .................................................................................................................................................. 3 Knowledge, Attitudes, and Practices............................................................................................................. 3 Introduction................................................................................................................................... 4 Context................................................................................................................................................................. 4 Shejeh Salam Program....................................................................................................................................... 4 Research Objectives.......................................................................................................................................... 5 Methodology .................................................................................................................................. 6 Sampling and Tools............................................................................................................................................ 6 Fieldwork OVerview ......................................................................................................................................... 7 Limitations, challenges and Considerations ................................................................................................. 8 Detailed Findings........................................................................................................................... 8 Community Peace Index................................................................................................................................... 8 CPI (a) ................................................................................................................................................................... 9 CPI (b).................................................................................................................................................................12 CPI (c).................................................................................................................................................................18 CPI (d).................................................................................................................................................................21 CPI (e).................................................................................................................................................................28 Custom Indicator 2..........................................................................................................................................31 Custom Indicator 3..........................................................................................................................................35 Custom Indicator 9..........................................................................................................................................39 Custom Indicator 19........................................................................................................................................42 Context Indicators ...................................................................................................................... 46 Annex 1: Indicator Performance Table .................................................................................... 50 Annex I1: Sampled Bomas ....................................................................................................... 101 Annex III: County conflicts....................................................................................................... 103 Annex IV: Overview of Local Conflict Dynamics ................................................................... 104 Annex V: Ranking of News Source Reliability ........................................................................ 111 Baseline Evaluation | Shejeh Salam Activity iii ACRONYMS CPI Community Peace Index CSO Civil Society Organization CSRF Conflict Sensitivity Resource Facility GBV Gender Based Violence IDP Internally Displaced Person INGO International Non-governmental Organization IO In-opposition IRC International Rescue Committee KAP Knowledge, attitudes, and practices LRA Lords Resistance Army NGO Non-governmental Organization PPS Proportionate to population size R-ARCSS Revitalized Agreement on the Resolution of Conflict In the Republic of South Sudan SPLA-IO Sudan People’s Liberation Movement In Opposition SSDM-Cobra Faction South Sudan Democratic Movement – Cobra Faction SSPDF South Sudan People’s Defence Forces UNDP United Nations Development Programme UNMISS United Nations Mission in South Sudan USAID US Agency for International Development WBEG Western Bahr el Ghazal Baseline Evaluation | Shejeh Salam Activity iv TABLE OF TABLES Table 1: Qualitative Data Sample ................................................................................................................................ 7 Table 2: Fieldwork Overview ....................................................................................................................................... 7 Table 3: Community Peace Index Scores .................................................................................................................. 9 Table 4: Community Peace Index (a) Scores .............................................................................................................9 Table 5: Community Peace Index (b) Scores ......................................................................................................... 12 Table 6: Community Peace Index (c) Scores .......................................................................................................... 18 Table 7: Community Peace Index (d) Scores ......................................................................................................... 21 Table 8: Community Peace Index (e) Scores ......................................................................................................... 27 Table 9: Custom Indicator 2 Scores ......................................................................................................................... 30 Table 10: Custom Indicator 3 Scores ....................................................................................................................... 33 Table 11: Custom Indicator 9 Scores ....................................................................................................................... 37 Table 12: Indicator 19a Scores .................................................................................................................................. 41 Table 13: Indicator 19b Scores .................................................................................................................................. 42 Table 14: Indicator 10c Scores .................................................................................................................................. 43 Table 15: Context Indicators, media access and exposure ................................................................................. 44 Table 16: Context Indicators, trusted sources....................................................................................................... 46 Baseline Evaluation | Shejeh Salam Activity 1 SUMMARY FINDINGS AND PROGRAMMATIC IMPLICATIONS INTRODUCTION This report presents baseline findings on the nature of conflict and peacebuilding in five regional clusters of South Sudan: Western Bahr el Ghazal, Unity, Upper Nile, Jonglei, and Eastern Equatoria. The report seeks to operationalize and measure values for key indicators in the targeted geographic locations which may be used to track changes over the program cycle. The evaluation employed a mixed-methods approach utilizing a community survey (inclusive of finalized indicators) and qualitative interviews with key informants and community members. Tools were designed to establish baseline values for indicators that will allow for comprehensive tracking throughout the activity’s lifecycle – through baseline, midline, and end-line. SAFETY AND SECURITY Indicators relating to safety and security include Custom Indicator 3 (percent of community members reporting that key conflicts have reduced in the past 18 months) and Community Peace Index (CPI) component (d) (percent of community members reporting that their personal safety and security has increased over the last 12 months). While Indicator 3 attempts to measure actual change in violence level – which occur at the community level and are not necessarily driven by individual experiences, CPI(d) seeks to understand personal perceptions of safety and may be vastly different within communities and even within households. Nonetheless, we find regional clusters perform similarly among the two indicators. Western Bahr el Ghazal (WBEG) ranks highest for both indicators, with the lowest reported levels of violent incidences and the highest perceptions of safety among respondents. For both indicators, WBEG is followed by Upper Nile and Southern Unity (respectively) while Jonglei and Eastern Equatoria appear to have the weakest safety and security conditions among the five regional clusters. In all locales, cattle raiding and revenge killings are the most prominent expression of conflict – confirmed by both qualitative and quantitative respondents. However, questions remain surrounding the frequency and violence of cattle raids. There is some evidence indicating cattle raids are relatively infrequent but are large in scale and high profile within counties, potential portraying the event as more prevalent than in reality. In the quantitative data, this is supported by the fact that very few bomas are confirmed by more than half of respondents to have experienced a cattle raid within the past year. Rather, many people are aware of the occurrence of cattle raids because of their high profile. Likewise, in several instances we find qualitative respondents referring to a similar incident to support their feelings that cattle raids are highly prevalent. We interpret this data to meant that we can have relatively high confidence that cattle raids are occurring, but some skepticism around the reported frequency of such events. Likewise, quantitative survey respondents offer varying perspectives as to whether cattle thefts in their county have become more or less violent over the past year. While 40% of respondents claim that raids have become more violent, a comparable share (47%) say the opposite and that they have become less violent. Qualitative data is equally mixed, with several respondents citing incidences in which cattle raids occurred and were resolved peacefully, while others report livestock theft associated with multiple casualties. Programmatic Implications: The data around safety and security indicators, which find substantial regional difference at the boma and county levels, have both programming and monitoring implications. On the programming side, the results caution against treating counties - let alone entire regional clusters - as homogeneous units and interventions should be localized to the greatest degree possible. Baseline Evaluation | Shejeh Salam Activity 2 From a monitoring perspective, the variation between bomas emphasizes the need to ensure longitudinal data is collected in the same locations to track progress. At times regional clusters do not appear to be meaningful units for any data trends and actually obscure county-level findings. Using the data at this level is likely to make it difficult to identify success stories and/or places where strategic programming adjustments may be appropriate. MECHANISMS AND CAPACITY FOR CONFLICT RESOLUTION, MITIGATION, AND PREVENTION In most counties, traditional leaders are the primary – and sometimes only – mechanism for conflict mitigation and/or resolution. As police or other state-backed actors are not present in many locals, the prevalence of customary leaders in maintaining community peace is necessary, if not always preferable. Although traditional leaders are generally respected and, based on qualitative data, often well-suited to respond to conflicts because of their knowledge of the community and local dynamics, they are at times insufficient. Without statutory backing, traditional leaders have limited capacity in responding to conflict once it has escalated to violence, as they are often unable to determine guilt between parties nor can they enforce any consequences on belligerents. While they are often best suited to prevent the escalation of violence, most respondents agree that the police or other state-backed actors need to be involved for large scale and/or violent conflicts. We also find that, where alternative mechanisms for conflict prevention and/or resolution are available, traditional elders are less popular. This again supports the idea that, while the traditional elders are respected, they are not necessarily the preferred peacebuilding actor but rather the de facto mechanism in many communities. Conflict resolution and mitigation is addressed through Community Peace Index component (b) (percent of counties with a functional mechanism for conflict prevention dialogue and conflict resolution) and Indicator 9 (the percent of local community members who believe that traditional authorities are playing their role in conflict prevention, resolution, or mitigation). Although Indicator CPI (b) scores are relatively similar at the cluster level, again we find that analysis at this level obscures clear county-level trends. In Eastern Equatoria, for example, we see vastly different scores in Budi, where confidence in conflict resolution mechanisms is low, and Kapoeta North, where it is extremely high. This is likely a reflection of the fact that the two locations have very different approaches to conflict resolution, with respondents in Budi almost exclusively relying on customary authorities while communities in Kapoeta North also take advantage of the police presence in the region to help resolve violent conflicts. A similarly stark difference emerges in Upper Nile, whereby respondents in Baliet report high levels of confidence in conflict mitigation actors while those in Ulang are far less so. Again, we find very different approaches to conflict resolution within the cluster as a majority of respondents in Baliet rely on the police for these services while those in Ulang go to customary authorities and religious leaders. Programmatic Implications: It will be important to tailor programming to local contexts at the most granular level possible, as regional clusters are often comprised of several highly varied county contexts. This may be particularly pertinent in ensuring interventions are conflict sensitive, as introducing or promoting new peacebuilding actors may threaten current power dynamics within communities as the current peace brokers or conflict resolution actors may feel they are losing ground in this area. In this sense, the program team should ensure they are cognizant of the specific actors of a given locale, as it will inevitably vary between locations. Baseline Evaluation | Shejeh Salam Activity 3 SOCIAL COHESION Custom Indicator 2 measures the percent of local community members reporting increased trust towards those that they have had a strained or past conflict within the past 18 months. In each county, the main conflicts and the primary actors or ethnic groups were identified. Respondents were asked to comment on the trends in violence and trust between these conflict actors. Community Peace Index component (e) also asks participants to comment on these relationships, measuring the extent to which these groups interact – through intermarriage, trade, and/or sharing resources. Eastern Equatoria stands out for having the lowest social cohesion scores for both indicators CPI (e) and Indicator 2. Although trade is somewhat common between rival groups, intergroup marriage and sharing of resource is rare. Qualitative respondents reiterate this message, saying that the region lacks key characteristics of peace such as the ability for communities to cooperate and coexist with each other. However, these indicators do not always correlate. In Southern Unity and WBEG, we see relatively high levels of interaction, without corresponding levels of trust and/or reduced tensions. This suggests that interventions which create avenues for increased interactions or dependencies between groups may not necessarily be associated with improved relationships. Programmatic Implications: The relationship between interactions between groups and stability or conflict levels is complex and should be treated with nuance. To illustrate it is likely that places where groups interact frequently necessarily – for example to share scarce resources - more opportunities for conflict may emerge. On the other hand, a political settlement may create a “cold peace” – or an environment of stability without cohesion. These distinctions are important as the guide the intent of social cohesion programming. The former scenario necessitates interventions aiming to improve the nature of interactions between groups, while the latter may benefit from interventions targeting the frequency of interactions. KNOWLEDGE, ATTITUDES, AND PRACTICES Program indicators address conflict-related knowledge, attitudes, and practices (KAP) among target communities through measuring knowledge of trauma (and its potential downstream effects) as well as propensity to utilize violence as a means of responding to conflict. Knowledge of trauma is extremely high in Southern Unity, Eastern Equatoria, and WBEG, with nearly all respondents demonstrating an adequate knowledge of trauma. Scores are lower in Jonglei and Upper Nile, where there is considerable variation at the county level. In Jonglei, for example, nearly all (98%) of respondents in Akobo demonstrate adequate knowledge of trauma while the same is true of just 4% in Duk. Pibor and Uror fall somewhere between these two poles, at 55% and 85%, respectively. As may be expected, a similar trend emerges for Indicator 19 (b), whereby Jonglei and Upper Nile lag behind the other regional clusters in their understanding of trauma and its relationship to violence and conflict. Finally, Community Peace Index component (c) measures the percent of community members who do not believe that violence is a viable way to resolve disputes. We find attitudes around this correspond to safety and security levels in a cluster, as defined in Indicator 3 and CPI (d). That is, those regional clusters which report the lowest perception of personal safety and highest incidence of violent conflict (Eastern Equatoria and Jonglei) also are the least likely to agree with the statement “The use of violence is never justified.” However, both quantitative and qualitative respondents in all regional clusters acknowledge that response to certain scenarios – such as livestock theft or impregnating a woman without paying dowry – may lead escalate to violence. Although respondents are not necessarily themselves condoning violence in all scenarios, it is revealing in its reflection of how some in the community think about – and respond to – conflict. Baseline Evaluation | Shejeh Salam Activity 4 INTRODUCTION CONTEXT This report presents baseline findings on the nature of conflict and peacebuilding in five regional clusters of South Sudan: Western Bahr el Ghazal, Unity, Upper Nile, Jonglei, and Eastern Equatoria. This research takes place in a pivotal time in South Sudan’s history as it seeks to emerge from the wake of December 2013’s violence and subsequent conflict in which more than 400,000 people have been killed and at least 4.2 million people have been displaced. The on-going Revitalized Agreement on the Resolution of Conflict in the Republic of South Sudan (R-ARCSS) is showing signs of ameliorating contemporary aspects of the war, which is shaped by more than 60-years of conflict in the country. Despite the R-ARCSS’s emergent success, there is a difficult road ahead. Challenges include: conflicts between political elites, which predate December 2013’s violence and can be traced for decades; conflicts between and within communities, including but not limited to land; conflicts between people and the state, including over changed state boundaries; distrust and fatigue in peace processes in the country; widespread trauma in the population; an oppressed and limited civil society; and a suppressed and weakened media environment. Given the range of threats and opportunities to a peaceful transition in South Sudan, this project seeks to support community-led peacebuilding and conflict mitigation initiatives. SHEJEH SALAM PROGRAM The U.S. Agency for International Development (USAID), in partnership with DT Global, is implementing a five-year, Shejeh Salam Activity in South Sudan. Using an integrated, multi-pronged approach, Shejeh Salam aims to lay the groundwork for a more stable and socially cohesive South Sudan, one that enables local actors to build community cohesion and promote peace processes; foster peaceful co-existence within and between communities; encourage civil society actors to advocate for peace and reconciliation; participate in civic processes; enable the provision of trauma awareness services to communities; and prevent misinformation by enabling access to accurate, fair and transparent information through different media channels. The proposed activity will take place in 13 (thirteen) counties in six (6) regional clusters – Juba, Southern Unity, Western Bahr el Ghazal, Upper Nile, Jonglei, and Eastern Equatoria – in South Sudan. The goal of the program is to “strengthen the foundation for a more self-reliant South Sudan” and is in alignment with USAID/Shejeh Salam Activity Objective 3 ‘Improved Social Cohesion in Targeted Areas’. To achieve this goal, USAID/Shejeh Salam has identified four objectives which they aim to achieve throughout the Activity’s lifecycle: I. Local actors are building cross line interdependence and intra-community cohesion to promote peace processes and peaceful co-existence II. Civil society actors are advocating for peace, justice, reconciliation, and reform; and participating in political and civic processes III. Key partners are providing trauma awareness services to communities IV. Print, radio, and other media are providing accurate, fair, and thorough information to mitigate the destructive impacts of rumors and misinformation Baseline Evaluation | Shejeh Salam Activity 5 RESEARCH OBJECTIVES This baseline report seeks to operationalize and measure values for key indicators in the targeted geographic locations which may be used to track changes over the program cycle. The indicators addressed, and their corresponding project objectives, are listed below. Objective Indicators An index made up of five indicators, each of which Goal: Improved social cohesion in targeted measures one of 5 specified outcomes of successful areas peacebuilding program. Objective 1: Local actors are building crossline interdependence and intracommunity cohesion to promote peace processes and peaceful co-existence Percent of local community members reporting increased trust towards those that they have had a strained past or current conflict with, in the past 18 months. Percent of community members reporting that key conflicts (namely cattle raid/theft, revenge killing, child abduction, or rape) have reduced in the past 18 months.1 Intermediate Result 1.1: Improved capacity of a wide range of local actors to engage constructively and successfully mitigated perceived conflict issues Percent of local community members who believe that traditional authorities are playing their role in conflict prevention, resolution, or mitigation Intermediate Result 1.2: Increased mutual interdependence resulting from infrastructure and livelihood projects Percent of community members reporting having initiated or increased interaction or relationship with members for whom they had a strained relationship or past conflict. Objective 3: Key partners are providing trauma awareness services to communities Percent of community members with awareness of trauma. 1 Mid-line and end-line surveys will cover Objective 4 (Shejeh Salam Peace Messaging), however this is not included in the baseline survey. Baseline Evaluation | Shejeh Salam Activity 6 METHODOLOGY SAMPLING AND TOOLS The evaluation employed a mixed-methods approach utilizing a community survey (inclusive of finalized indicators) and qualitative interviews with key informants and community members. Tools were designed to establish baseline values for indicators that will allow for comprehensive tracking throughout the Activity’s lifecycle – through baseline, midline, and end-line. The sampling approach to the quantitative data collection tool was a multi-stage, stratified cluster sample. The first stage of selection, the 13 target counties, was determined based on the need to capture project target locations. The second stage consists of selecting bomas from within counties. Within each stratum, we selected bomas randomly with probability proportionate to population size (PPS) sampling.2 Once clusters are assigned to bomas, the final sampling stage consists of random household selection. This is achieved through the random walk approach, in which we identify the populated area of the boma and selecting random starting point(s). Within households, selection of the member to be interviewed was randomized.3 It is important to note that Shejeh Salam Activity does not intend to operate in all bomas of a county nor will the exposure to activities be equal among all targeted bomas. As such, boma sampling prioritized “high intensity” bomas for selection, only including medium or low intensity bomas when the sampling required it. Where interventions in each county are only being implemented in one or two bomas in total, we include “Low Intensity” implementation locations to ensure sufficient sample size. Annex 1 summarizes the bomas targeted in the final sample. We utilize a cluster size of ten interviews, meaning that a boma will be assigned ten interviews each time it is selected using PPS sampling.4 The sample target specified180 clusters, for a total of 1,800 quantitative respondents. The final realized sample varied somewhat, as enumerators occasionally overcollected. And, in one instance, a cluster of 6 was collected, falling short of the target of 10 respondents per cluster. Ultimately, a sample of 1,848 respondents was achieved. The objective of the quantitative tool was to collect data which would measure and address the Shejeh Salam Activity indicators. The data from the quantitative tool is central to this report’s analysis. To complement the quantitative data, qualitative discussion guides were also developed. The purpose of the qualitative interviews was to capture more detailed, free-form evidence regarding changes in community perception surrounding trust, conflict reduction (cattle-raiding, revenge killing, theft), the role of local authorities, social cohesion, and receptivity to Shejeh Salam peace education and media messaging. Interviews will target knowledgeable and active individuals in each community, who have a wide perspective on issues within the communities, events, and opinions. The following table provides an overview of the respondents interviewed in each target county. 2 Some adjustments were made to the random selection based on the accessibility and security of bomas. 3 As response biases in the data emerged, enumerators were instructed to adjust household member selection to try and capture underrepresented populations. For example, in areas where few youth male respondents were available, enumerators would employ a “take-all” approach to that group, meaning if a youth male was available in a household, they would select them for participation instead of randomizing the participant. 4 Recall that PPS sampling employs sampling with replacement, such that bomas can be selected multiple times. Baseline Evaluation | Shejeh Salam Activity 7 Table 1: Qualitative Data Sample Regional Cluster County Amount Groups/Respondents Totals Southern Unity Mayendit 2 Community members with at least 50% female and 50% youth 6 Leer 2 Panyijar 2 Eastern Equatoria Kapoeta North Budi 3 3 Community members with at least 50% female and 50% youth 6 Western Bahr el Wau 3 Community members with at 6 Ghazal Jur River 3 least 50% female and 50% youth Jonglei Akobo Duk 2 1 Community members with at least 50% female and 50% youth 6 Pibor 1 Uror 2 Upper Nile Baliet 3 Community members with at least 50% female and 50% youth 6 Ulang 3 30 FIELDWORK OVERVIEW Data collection took place between 7 May and 2 June 2021, with county-specific dates outlined in the table below. Table 2: Fieldwork Overview County Start Date End Date # Baliet 7-May 22-May 187 Duk 11-May 17-May 92 Leer 12-May 23-May 128 Mayendit 13-May 16-May 120 Panyijar 13-May 21-May 120 Pibor 15-May 22-May 91 Budi 18-May 22-May 181 Uror 19-May 24-May 84 Akobo 20-May 24-May 91 Ulang 20-May 27-May 211 Baseline Evaluation | Shejeh Salam Activity 8 Jur river 21-May 25-May 180 Kapoeta North 25-May 28-May 183 Wau 25-May 2-Jun 180 LIMITATIONS, CHALLENGES AND CONSIDERATIONS Accessibility: Security and accessibility issues challenged the data collection efforts in multiple locations. In these cases, replacement bomas were selected. Although targets were ultimately met, this posed somewhat of a limitation in that respondents a were more concentrated in some geographies than initially intended. DETAILED FINDINGS In this section we present the main findings for each of the evaluated indicators. The section is organized by Indicator, then disaggregated by regional cluster. For each Indicator, we begin by describing the method through which the indicator was measured. We then go on to briefly discuss our confidence in the indicator as a measure and, where relevant, identify potential limitations of the measure. Within each regional cluster, the structure of the analysis varies somewhat, depending on the findings. In other words, we describe only what is interesting or significant, but do not describe all trends (or lack thereof). In Annex X, we report all indicator scores for each regional cluster, disaggregated by gender, age group, and ethnic group. COMMUNITY PEACE INDEX The Community Peace Index is made up of five indicators, each measuring a specified outcome of a successful peacebuilding program. The five indicators are described as follows: a. Percent of community members reporting that they are part of, or aware of, a local (community-led) conflict prevention, resolution, or mitigation mechanism that was developed within the last 12 months in targeted areas.5 b. Percent of counties with a functional mechanism for conflict prevention dialogue and conflict resolution. c. Percent of community members who do not believe that violence is a viable way to resolve disputes in targeted areas. d. Percent of community members reporting that their personal safety and security has increased over the last 12 months, in targeted areas. e. Percent of community members reporting having initiated or increased interaction or relationships (in the past 18 months) with members for whom they had strained relationships or past conflict, in targeted areas. The index seeks to measure stability indicators which measure progress in peacebuilding and are presumed to correlate. In other words, we assume that there is some relationship between these indicators and that they are likely to rise and fall together, dependent on the success (or lack thereof) of peacebuilding in the location. The reliability of the indicator is measured using Cronbach’s Alpha, which tests the inter-item covariance of the index’s five indicators. Using this test across all five indicators, we find low overall or composite correlation (0.52) between the presence of local community-led conflict prevention mechanisms in surveyed locations in South Sudan and respondents’ awareness of them or 5 This refers only to those started by the community or community leaders and excludes initiatives driven by external actors, such as INGOs. Baseline Evaluation | Shejeh Salam Activity 9 changed attitudes towards violence and conflict. This suggests that these variables do not operate as an index, in that the different components do not appear to increase or decrease together. The low composite Alpha score of the index may be understood in several ways. The first is the extent to which the five components of the index are related – or associated in anticipated ways. For example, Item a identifies locations where new conflict prevention, resolution, or mitigation mechanism has been introduced; meanwhile, Item d concurrently assesses improved safety and security. Although the introduction of a conflict mitigation mechanism may be assumed to result in improved safety on a longer timescale, this index is a short-term snapshot. Significantly, when viewed through the latter lens, the contemporaneous measurement of security and the introduction of new conflict mitigation mechanisms are likely to have an inverse relationship. In other words, new conflict mitigation mechanisms are expected to be introduced where safety is identified as deteriorating. Conversely, the introduction of new conflict-related mechanisms where security is improving is unlikely. As such, we report aggregate Index scores, but in our discussion, we treat the five components as individual indicators. The baseline survey yielded an overall CPI score of 55%, which varies slightly by state, gender or ethnic group. WBEG, Upper Nile, and Southern Unity states have the highest CPI values, at 69%, 60%, and 59%, respectively, which are slightly above average. Eastern Equatoria and Jonglei fall below the average at 38% and 48%, respectively. While the Dinka, Nuer, Luos and Toposa also rank above average, the Buya and Murle present the lowest CPI values at 21% and 40%, respectively. Women tend to demonstrate higher scores than men. However, the following sections will illustrate the relevant trends in relation to factors such as state, district, gender, youth, or ethnic group for each CPI indicator, separately. Table 3: Community Peace Index Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Total Cluster Average Southern Unity 0.54 0.53 0.54 0.62 0.60 0.61 0.58 Eastern Equatoria 0.37 0.39 0.38 0.35 0.45 0.38 0.38 Western Bahr el Ghazal 0.69 0.69 0.69 0.70 0.62 0.67 0.68 Jonglei 0.43 0.45 0.44 0.45 0.43 0.44 0.44 Upper Nile 0.59 0.54 0.55 0.59 0.61 0.60 0.58 CPI (A) Definition: Percent of community members reporting that they (respondent or other household member) are part of, or aware of, a local (community-led) conflict prevention, resolution, or mitigation mechanism that was developed within the last 12 months in USAID/Shejeh Salam targeted areas. Operationalizing the indicator: The indicator measures the percent of community members who answer “yes” to the following question: “As far as you know, over the past year have there been any new Baseline Evaluation | Shejeh Salam Activity 10 peacebuilding systems or projects in your community? These can include people or groups who try and prevent conflict, resolve disputes, or maintain peace in the community.” We included only those who referenced ones initiated by community members or community leaders. Table 4: Community Peace Index (a) Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Total Cluster Average Southern Unity 0.06 0.08 0.07 0.18 0.20 0.19 0.14 Eastern Equatoria 0.06 0.07 0.07 0.06 0.15 0.08 0.08 Western Bahr el Ghazal 0.33 0.41 0.37 0.29 0.30 0.29 0.32 Jonglei 0.12 0.07 0.09 0.11 0.05 0.08 0.08 Upper Nile 0.28 0.20 0.22 0.27 0.34 0.30 0.27 SOUTHERN UNITY In all three surveyed counties – Mayendit, Leer, and Panyijar – few respondents have heard about a new peacebuilding project in the community. Among the three counties, community members in Panyijar were most likely to have heard about a peacebuilding project – though when pressed for details, few respondents could describe the project in any detail. A few respondents cited an International Rescue Committee (IRC)-led gender-based violence (GBV) initiative, and over a dozen respondents in the county made vague reference to “Mediators”. However, most said they “don’t know” and, when asked about the actors behind the peacebuilding project, there is a general indecisiveness regarding whether community leaders, a local civil society organization (CSO), the county government, or an international non-governmental organization (INGO) were individually or collectively behind it. As such, although residents of Panyijar are somewhat more likely to have heard about a community-driven peacebuilding project than those in Mayendit and Leer, information remains low and generally unclear. Alternatively, while a smaller share of respondents in Mayendit reported the existence of a peacebuilding project in their community, there was considerably more clarity among those who are aware of such a project. Nearly all reference two national peacebuilding and development organizations, that are the Dialogue and Research Institute and Hope Restoration. Most attribute this project to local or community leadership and/or the county government. In Leer, just three people are aware of such a project. Across the three counties, knowledge of peacebuilding projects does not vary by age group nor gender. JONGLEI Indicator CPI (a) scores vary considerably in Jonglei counties. Though most (59%) respondents in Akobo County report knowledge of a peacebuilding project in their community, this is only true for a minority in other counties in Jonglei, namely Duk (21%), Uror (17%), and Pibor (3%). Most respondents (58%) attribute the program to an INGO, though a substantial share also mentions local or community leaders (22%) and/or local CSOs (12%) as actors involved in the program. In contrast to other parts of South Sudan, Jonglei has been host to many community driven peace agreements that have been internationally supported, including by UNMISS in Akobo and UNDP support to community peace dialogues in Duk County. Baseline Evaluation | Shejeh Salam Activity 11 Women are more likely to be aware of community peacebuilding projects than men, even when controlling for age group and county. This finding is somewhat unexpected given the larger role that men typically play in public peace processes than women and suggests that women potentially informally contribute to these processes. However, one may argue that peacebuilding projects could be understood in a much broader sense than as just a peace process and be associated with livelihood and development projects. In that case, it is indeed likely that women rather than men would know about them, as these initiatives tend to involve more women. UPPER NILE A slight majority (52%) of respondents in Ulang County were aware of peacebuilding projects in their community, compared to only 31% of respondents in Baliet county. The increased prevalence of the peacebuilding awareness might be attributable to the way in which Ulang County was more closely ensnared in December 2013 related Upper Nile conflict than parts of Baliet County were. Most respondents in Baliet County indicated that they were aware of a National Dialogue effort (government sponsored) in their locality. Most of these respondents identified state actors (Police or Military, 42%) and customary/traditional leaders (27%) as being involved in the programs. EASTERN EQUATORIA In Kapoeta North, 17% of respondents were aware of a peacebuilding project in their community, compared to only a few respondents in Budi. This discrepancy might be explained by the rounds of negotiations and community consultations that have taken place in the county to resolve local conflicts and the way in which Budi has been comparatively less impacted by national conflict than Kapoeta North’s population has been. Those who were aware of such efforts in Kapoeta North referred to “Reconciliation”; indicating that the involved actors were either members of the military (38%) or customary authorities/traditional leaders (47%). These responses might speak to the prevalence of negotiations and community consultations that have taken place to resolve cattle raiding and intercommunal clashes in the county. Among qualitative interviews, respondents are aware of primarily health and education related interventions in the community, implemented by INGOs. In fact, respondents attribute most positive changes and/or developments in the community to INGO activity, with little reference to local or national actors. Rather, participants lament the ineffectiveness of the government in improving local conditions – including failing to provide infrastructure, school operating expenses, and/or health care – while also praising NGOs for their ability to step in on such matters. WESTERN BAHR EL GHAZAL Respondents in Jur River County were more likely (40%) to be aware of peacebuilding projects in their community relative to respondents in Wau county (30%). Among respondents who were aware of peacebuilding projects in Wau, references were made to several initiatives including: “Defy Hate/Defy Hate Speech”, “Sports for Peace (or some variation thereof), and “Peace Group (although it is unclear if this is being used a generic term for any number of peacebuilding initiatives that were part of the TriState Conference). While some respondents in Jur River recognized “Unity and Peaceful Coexistence” as a peacebuilding initiative in their community, respondents here tended to identify the parties, associated with the peacebuilding activities (e.g. spiritual leaders, NGOs, Chief Groups, etc.) rather than the initiative itself. In both Jur River and Wau, respondents tended to associate peacebuilding projects with county or local level organizations rather than the National Government or International NGOs. Only 15% of the respondents in Wau and 20% of the respondents in Jur River associated the peacebuilding activities with either the National Government or international NGOs. Instead, “local or community leadership (33%) Baseline Evaluation | Shejeh Salam Activity 12 and the county government (~27%) were more likely to be identified as the actors responsible for peacebuilding activities. Several key informants in WBEG are aware of, or have participated in, peacebuilding initiatives in the region. One, a county official in Jur River, mediated a dispute between two communities in Jur River County and people from Warap following a cattle raid. After several meetings with chiefs and women from the communities, according to the respondent, a peaceful resolution was met and the youth offenders ultimately returned the cattle to the owners. Another key informant, a CSO representative, has also been involved in resolving disputes related to cattle theft. The respondent’s organization is part of the Peace and Reconciliation Working Group of WBEG and facilitate peace dialogues in the community as needed. Although these cases are anecdotal, it is worth noting the presence and apparent ongoing activity of local peacebuilding activities in the cluster. CPI (B) Definition: Percent of USAID focus counties (13) with a functional mechanism for conflict prevention dialogue and conflict resolution. Operationalizing the indicator: Respondents were asked to consider a scenario where someone in their family experienced livestock theft. After identifying the person or actor they would contact to prevent the issue from escalating into violence, respondents measured their confidence in their ability to successfully prevent violence. Respondents were then asked to imagine that the scenario had become violent and again identified the person or actor they would contact to resolve the conflict. Finally, respondents were asked to measure their confidence in their ability to successfully resolve the conflict after it had already escalated into violence. To measure indicator CPI (b), we take the average confidence levels – (1) confidence in their ability to prevent conflict from escalating into violence, and (2) confidence in their ability to resolve conflict after it has become violent. Using Cronbach’s Alpha test, we find a high reliability coefficient for these two variables (0.84), indicating we can have some confidence in the use of these two variables to measure a single indicator. Table 5: Community Peace Index (b) Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Cluster Average Southern Unity 0.72 0.75 0.74 0.80 0.78 0.79 0.77 Eastern Equatoria 0.77 0.70 0.74 0.63 0.75 0.67 0.70 Western Bahr el Ghazal 0.87 0.81 0.84 0.88 0.81 0.86 0.85 Jonglei 0.79 0.80 0.79 0.76 0.72 0.74 0.76 Upper Nile 0.79 0.73 0.75 0.73 0.73 0.73 0.74 Preferred Actor For Conflict Prevention, By County Baseline Evaluation | Shejeh Salam Activity 13 Figure 1: CPI(b) Preferred Actor for Conflict Prevention, by County SOUTHERN UNITY Respondents in Southern Unity report general confidence in their ability to both prevent and resolve conflicts – suggesting the presence of functional mechanisms for conflict prevention dialogue and conflict resolution. There is some variation across counties – both in the degrees of confidence as well as the conflict prevention and resolution mechanisms. For instance, in Mayendit, the county within Southern Unity with the lowest score for this indicator, respondents demonstrate a substantial dependence on customary authorities for both the prevention of violence and resolution of conflicts. As many as 80% of respondents in the county say that they would turn to customary authorities to prevent a conflict from escalating and 74% would turn to customary authorities to resolve a violent conflict. The military are the next most cited actors that community members would turn to for both prevention and resolution. These patterns likely speak to the impact that negotiations and community consultations have had in their attempt to resolve conflicts in Mayendit. Customary authorities are also the preferred mediators for more than 50% of respondents in Leer and Panyijar, though they are preferred by less decisive margins than reported in Mayendit. Police are favored by more than 40% of respondents in Panyijar and just over 20% of respondents in Leer. Military personnel, meanwhile, are rarely selected as choice mediators for either the prevention or resolution of conflict in either Panyijar or Leer. At the cluster level, women are generally more confident in the conflict prevention and resolution mechanisms available to them, though this difference in minor.6 Despite this difference, there is little variation between men and women in their preferred conflict prevention and/or resolution mechanisms. 85 109 58 136 74 84 Baseline Evaluation | Shejeh Salam Activity 14 86 96 Key informants speak more favorably of customary authorities than of police, though there is some skepticism over the degree to which either is effective in maintaining peace. Generally, customary authorities are viewed as closer to the community and best suited to respond to most issues. For more serious infractions, although the customary authorities will be included, community members are likely to seek support from police as well. JONGLEI Confidence in conflict prevention and resolution mechanisms in Jonglei is relatively high at the cluster level, though there is substantial variation between counties. Indicator CPI (b) scores are lowest in Akobo and Pibor, where less than a third of respondents are “very confident” in the existing mechanisms to either prevent or resolve conflict. These shares are much higher in Duk and Uror where 80% and 77%, respectively, are very confident in the conflict prevention mechanisms and a similar share (77% in Duk and 86% in Uror) say the same of the conflict resolution mechanisms. The increased trust in conflict mechanisms in Duk is likely the outcome of community-driven efforts in both counties to resist violence and increase Nuer-Dinka cooperation. Indeed, in 2014 the International Crisis Group 6 The difference is statistically significant. notes that these were the strongest intercommunal relations between these two groups shortly after the conflict emerged. There is also considerable variation in the preferred actors or mechanisms for conflict prevention and resolution. In Akobo, nearly all respondents – for both prevention and resolution of conflict – utilize customary authorities or the county commissioner – with a roughly 50/50 split between the two. More specifically, most respondents in the county have a single preferred mechanism (traditional authorities) that they turn to for both the prevention and resolution of conflict. Similar dynamics are found in Uror, where most people would turn to customary authorities for both the prevention and resolution of conflict. Pibor echoed Uror in this sense, as most people would turn to customary authorities (more than 65%) or the police (approximately 20%). This too likely speaks to the role customary authorities had in improving Nuer-Dinka relations in the region and from the region into neighboring counties. However, in Uror county, a meaningful share of those who said they would turn to customary authorities to prevent the escalation of conflict said that they would instead go to the police once a conflict became violent. In fact, while just a quarter (26%) of respondents in Uror would seek assistance from the police in the prevention of conflict, nearly half (44%) would turn to the police for support in resolving a conflict once it became violent. Duk is somewhat unique to the cluster, as the majority (75%) of respondents in the county would turn to the police to prevent conflict from becoming violent and just 20% would rely on customary Baseline Evaluation | Shejeh Salam Activity 15 authorities. Once a conflict becomes violent, customary authorities are somewhat more popular (25%) and about 11% would seek support from the county commissioner. Confidence in Conflict Prevention Mechanisms: Jonglei 93 97 69 121 83 81 121 43 No significant trends emerge in analyzing Indicator CPI (b) by gender and age group. While there are differences between ethnic groups, ethnic groups form at the county level and therefore the extent to which trends can be attributed to ethnic group versus social, political, and economic conditions within a county is unclear. Nonetheless, among the three main ethnic groups surveyed in Jonglei, respondents who identified as Dinka are substantially more confident in the effectiveness of both conflict prevention and resolution mechanisms available to them than Nuer and Murle identifying respondents in the cluster. There is also a stronger tendency for Dinka identifying respondents to rely on police, rather than traditional authorities, as is preferred by Nuer and Murle. On the other hand, respondents who identified as Nuer are most likely to seek assistance from the county commissioner than either of the other main ethnic groups in Jonglei. Respondents who identified as Murle more consistently rely on traditional authorities for both conflict prevention and resolution. UPPER NILE Respondents in Upper Nile cluster from both sampled counties (Baliet and Ulang) expressed medium to high amounts of confidence in conflict prevention mechanisms in their locality. Again, this observation is likely a reflection of the relative success of community driven initiatives in these counties to forge ‘islands of peace’ during some of the most difficult waves of post-December 2013 waves of violence. When examined at the county level, respondents differed greatly on which institutions they would approach to prevent a conflict from turning violent. In Baliet, most indicated that they would turn to the police (52%) or to customary authorities (35%). In Ulang, the opposite is true. Most respondents (78%) indicated they would turn to customary authorities, and only 3% to police. Respondents did not change their preferences when asked who they would approach should the conflict have already turned violent. Baseline Evaluation | Shejeh Salam Activity 16 Confidence in Conflict Prevention Mechanisms: Upper Nile Qualitative respondents elaborated on the benefits and drawbacks of both the police and customary authorities and conflict mitigation actors. Generally, key informants revealed a preference for customary authorities who are closer to the community and have a better understanding of local dynamics. Traditional elders are seen as effective in advising peaceful ways to prevent or resolve conflicts. However, their role does not include determining guilt or adjudicate the winners or losers in disputes. Such duties are reserved for the state judicial system and the police. However, qualitative respondents say that some community members avoid involving the police for a number of reasons. First is the risk of escalation, as several respondents insinuate a fear of police themselves. However, the police are also susceptible to bribes and/or nepotism in their behavior – according to several key informants. The police may also extort money from community members seeking their services in resolving a dispute. “…if you have a problem you better go to the customary authorities, who have no gun.” CSO Representative, Baliet, Upper Nile EASTERN EQUATORIA Respondents in Eastern Equatoria expressed comparable levels of confidence in conflict prevention and resolution mechanisms relative to other regional clusters. Respondents residing in both Budi and Kapoeta north preferred approaching customary authorities/traditional leaders, although this preference was especially stark in Budi where half (50%) of the respondents indicated that they would turn to customary authorities/traditional leaders in the case of a dispute they felt may turn violent. 17% of respondents in Kapoeta North indicated that they would approach the police, while not a single respondent in Budi indicated that they felt the police were an appropriate resource for dealing with an impending conflict. Respondents did not significantly change their preferences when asked if they would approach the same conflict resolution body if a conflict had already turned violent. Age, gender, and ethnicity likewise had minimal effect on which preference a person had. 83 60 56 87 127 95 94 128 Baseline Evaluation | Shejeh Salam Activity 17 Confidence in Conflict Prevention Mechanisms: Eastern Equatoria Consistent with the other regional clusters, this is not necessarily reflective of preference for traditional leaders, but rather a consequence of a lack of options for dispute resolution actors. Qualitative respondents elaborate on this, explaining that in some communities, the police are not present and one must travel a long distance (more than a day, by foot, according to one respondent) to reach the nearest police station. The lack of police presence is broadly viewed as a problem and the traditional leaders are described as insufficient in meeting the needs of communities in the regional cluster. Even amongst leaders in the community, placing sole responsibility upon traditional authorities for conflict prevention or de-escalation is viewed as problematic. This is illustrated in the below quote from a CSO representative from Eastern Equatoria. “In the absence of the government, the chief becomes the only solution to the problems who can handles some mere disputes among the community like; loans, fighting, land issues and other forms of social disagreement.” CSO Representative, Eastern Equatoria WESTERN BAHR EL GHAZAL Of the five regional clusters, respondents in WBEG report the highest confidence in their conflict prevention and resolution mechanisms, which might speak to the successful impact that communitydriven initiatives such as the tri-state conference had in the region. In both Wau and Jur River, police are the preferred actor for support in both preventing violence and resolving conflict once it has become violent. This contrasts with other regions, where customary authorities tend to be respondents’ preferred actors. In WBEG, just 16% of respondents would turn to customary authorities to prevent a conflict from becoming violent and 18% would turn to them to help resolve a conflict once it became violent. While female respondents are slightly more inclined to turn to police than male respondents, the difference is minor, and the police remain the clear preference among both groups. However, age group appears to be a more reliable determinant of a respondent’s view of the police than gender; youth are significantly more likely to turn to the police than non-youth.7 On the other hand, non-youth respondents in WBEG are more likely to turn to customary leaders than their youth counterparts.8 Indicator scores – or a respondent’s confidence in these actors – also varies among demographic groups. A respondent’s gender, age group, and county are statistically significant predictors of indicator 26 41 22 45 30 62 43 49 88 117 98 107 Baseline Evaluation | Shejeh Salam Activity 18 CPI (b) scores. At the regional cluster level, female respondents report higher confidence levels than male respondents and youth higher than non-youth. At the same time, respondents in Wau county report higher levels of confidence in conflict prevention and resolution mechanisms than those in Jur River. Confidence in Conflict Prevention Mechanisms: WBEG 84 182 98 168 35 30 53 48 It is also worth noting some difference between ethnic groups in the region. In Jur River, respondents mainly identify as either Luo or Dinka. In constraining indicator CPI (b) analysis to these two ethnic groups in Jur River, we find Dinka respondents score significantly higher – or reporting higher confidence in conflict prevention and resolution mechanisms in the county.9 In Wau, respondent ethnic groups vary more, and no clear trends emerge based on this demographic characteristic. CPI (C) Definition: Percent of community members who do not believe that violence is a viable way to resolve disputes in Shejeh Salam targeted areas. Operationalizing the indicator: Respondents are presented with two statements and asked to report the extent to which they agree or disagree with the statement. The statements read as follows: • Statement 1: The use of violence is never justified. • Statement 2: In this country, it is sometimes necessary to use violence in support of a just cause. Covariance between responses to these two statements is extremely low, such that we cannot reliably use the two variables together as an index. Because the first statement is a more straightforward interpretation of the indicator (defined as the share of community members who do not believe that 7 The difference is statistically significant when controlling for gender. 8 The difference is statistically significant when controlling for gender. Baseline Evaluation | Shejeh Salam Activity 19 9 The difference is statistically significant when controlling for gender. violence is a viable way to resolve disputes, in Shejeh Salam targeted areas) we take the share of those who “Strongly Agree” with Statement 1 as a measure of indicator CPI (c). Table 6: Community Peace Index (c) Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female State Average Southern Unity 55% 57% 56% 74% 67% 71% 66% Eastern Equatoria 5% 21% 13% 12% 29% 17% 15% Western Bahr el Ghazal 47% 38% 43% 51% 24% 42% 43% Jonglei 24% 22% 23% 24% 24% 24% 24% Upper Nile 37% 34% 35% 55% 53% 54% 46% Response to the statement: "The use of violence is never justified." SOUTHERN UNITY At the cluster level, 66% of respondents strongly agree that violence is never justified. This is highest in Panyijar, where the share of respondents who strongly agree is 96%. Though few people disagree in any county, a substantial share of respondents in both Leer and Mayendit said they only “somewhat agree” and about half (46% in Leer and 58% in Mayendit) strongly agree with the statement. Women were much more likely to agree with the statement; while 72% of female respondents strongly agree with the statement, this is true of just 57% of male respondents in Southern Unity.6 There is no discernible difference between youth and non-youth respondents in the cluster. JONGLEI Compared to other regional clusters, Jonglei scores relatively low on Indicator CPI (c). However, while fewer respondents “strongly agree” with the statement than in most other clusters (scoring higher only 6 The difference is statistically significant, even when controlling for county. 6 224 129 76 137 12 57 48 74 22 105 28 21 122 46 243 55 154 85 185 Baseline Evaluation | Shejeh Salam Activity 20 than Eastern Equatoria), relatively few respondents “strongly disagree” with the statement. In fact, there is relatively even distribution of responses across the Likert scale. Trends are equally fuzzy at the county level in Jonglei. Within the clusters, scores are lowest in Akobo, where just 3% of respondents strongly agree with the statement (though at the same time, 0% strongly disagree with the statement). In Uror, although roughly a third (32%) strongly agree with the statement, more than half of respondents in the county disagree, to some extent. Nearly all respondents in Pibor agree to some extent and the county has the highest share of “strongly agree” respondents. However, it is difficult to find meaningful relationships between Indicator CPI (c) scores and demographic or environmental variables. There is not a significant relationship between the indicator and: age group, ethnic group, nor gender. Interestingly, the only significant correlate with the indicator is whether respondents have moved between bomas or counties over the past year. While controlling for county, gender, and age group, respondents who said “yes” they have moved between bomas over the past year are less likely to agree with the statement.7 Many key informants associate violence in the region with alcohol abuse. There are also mentions of cattle raids as triggers of violent conflict – particularly during the rainy season. Revenge attacks also occur, as well as disputes resulting from early pregnancies. The range of triggers for violence described by key informants suggests that, at least among some in the community, violence is viewed as a viable response to conflict. “Mostly the youth who drinks alcohol sometimes causes problems and sometimes the fight becomes violent” Community Leader, Uror, Jonglei UPPER NILE Respondents in Upper Nile are split along this indicator, with 46% indicating that they “strongly agree” that violence is never justified and 34% answering that they strongly disagreed. These differences become clearer at the county level, where respondents in Ulang county are much more likely to disagree than respondents in Baliet. However, both counties have significant populations that both strongly disagree and strongly agree. In this level, it appears that correlation with opinion lies mostly with geography and the security situation at the local level. Discerning clear trends based on demographic indicators in Upper Nile state was difficult, although gender was a clear predictor; with women being twice as likely to indicate agreement than men. Age did not have a discernable effect on opinion within the Upper Nile state. Although key informants were not asked to respond to the same statement as quantitative respondents, participants shared their thoughts on some of the drivers of violent conflict in the region. A variety of catalysts for violence were presented, including retaliation following theft, failure to repay a loan, and failing to pay dowry after impregnating a woman. Several respondents offer ways in which conflicts can 7 The difference is statistically significant. Baseline Evaluation | Shejeh Salam Activity 21 be resolved prior to the escalation into violence, such as meetings with chiefs (Omda). Nonetheless, the variety of scenarios presented which may lead to violence – though not necessarily reflective of the respondents’ views themselves – does suggest a level of acceptance within the communities in responding to conflict with violence. EASTERN EQUATORIA At the cluster level, 61% of respondents strongly disagree that violence is never justified. This is a stark difference from other clusters, where most respondents indicated that they agree that violence is never justified. Respondents from Budi County were most likely to exhibit strong disagreement, with 70% of voicing that they thought violence was justified in some circumstances, compared to 52% in Kapoeta North. Budi is one of the most conflict-prone locations in the state for certain types of violence – 98% of respondents indicated that they knew someone who had been affected by a cattle raid and 97% knew of someone who had been affected by a revenge attack. Given Budi’s high correlation with these types of violence, it is apparent that security is weak and residents then justify the use of violence in specific contexts. “Two weeks ago, Buya attacked one of the cattle camps and went away with some heads of cattle in a bloodshed operation but till now our youth have not responded yet.” Key Informant, Eastern Equatoria WESTERN BAHR EL GHAZAL Although WBEG scores relatively high on indicator CPI (c) compared to other regional cluster, there is meaningful variation between sub-groups. The first is at the county level, where respondents in Wau are more likely to agree with the statement than those in Jur River.8 In fact, more than half of respondents in Jur River disagree (to varying extents) with the statement: “The use of violence is never justified.” There are also considerable differences in opinion between youth and non-youth respondents in the region, in that youth are more likely to agree with the statement than non-youth.9 However, we do not see meaningful differences between male and female respondents nor between ethnic groups. CPI (D) Definition: Percent of community members reporting that their personal safety and security has increased over the last 12 months, in Shejeh Salam targeted areas. 8 The difference is statistically significant. 9 The difference is statistically significant when controlling for gender and county. Baseline Evaluation | Shejeh Salam Activity 22 Operationalizing the indicator: This indicator measures the percent of community members reporting that their personal safety and security has increased over the last 12 months, in Shejeh Salam targeted areas. This does not provide absolute levels of safety or security. As such, the indicator should not be used to compare security levels in different locations. Rather, it should be used within a location to conduct temporal analysis of change in trends. As this is the baseline, we are not measuring change in safety perceptions, but rather setting a benchmark for safety to be compared against midline and endline evaluations. Table 7: Community Peace Index (d) Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Total State Average Southern Unity 0.52 0.46 0.48 0.52 0.48 0.50 0.50 Eastern Equatoria 0.36 0.37 0.37 0.33 0.44 0.37 0.37 Western Bahr el Ghazal 0.88 0.88 0.88 0.87 0.86 0.87 0.87 Jonglei 0.39 0.44 0.43 0.46 0.47 0.46 0.45 Upper Nile 0.68 0.57 0.60 0.55 0.58 0.57 0.58 SOUTHERN UNITY Perceptions of safety in Southern Unity are highest in Panyijar, followed by Leer, and then Mayendit. This pattern is reflected both in the indicator question (“How would you describe the safety of your boma?”) and in follow up questions designed to test community safety. Respondents were asked whether they would feel comfortable with a female family member walking around alone at night. In Panyijar, 41% of community members said they would be comfortable with this, while just 19% in Leer said the same and just 2% in Mayendit. Youth respondents in Southern Unity score higher on indicator Baseline Evaluation | Shejeh Salam Activity 23 CPI (d) than non-youth.10 On the other hand, there is no notable difference between male and female respondents – apart from Mayendit county, where women score significantly lower than men. Although most key informants generally describe the region as safe, roads are identified as potentially insecure, depending on the connecting regions. Several qualitative respondents in Southern Unity mention the proliferation of arms among youth. This is attributed by one key informant to the weakness of the state and policing bodies creating a culture of impunity. However, another feels that it is because they fear for their lives. A third says that youth arm themselves to participate in cattle raids. Regardless, there appears to be some consensus among qualitative respondents for government action to disarm youths as a means of improving community safety. “They know the government is very disorganized and that nobody will follow you.” Female community member, Leer, Southern Unity 10 The difference is statistically significant. Baseline Evaluation | Shejeh Salam Activity 24 Of the five clusters, Jonglei has the second lowest score for Indicator CPI (d). Geography appears to be the most meaningful predicator of the Indicator. At the county level, Duk and Pibor score significantly lower than Akobo and Uror. Within Duk and Akobo counties, variation between bomas does not vary substantially; in Akobo all bomas report relatively moderate levels of safety while in Duk bomas consistent report poor safety conditions. This is not the case with Pibor county, where there is far more variation in perceptions of safety between bomas. This is particularly true in Lawol and Wunkok bomas, where perceptions of safety are significantly lower – both in terms of statistical significance and in magnitude of difference. Variation in Uror is somewhat more tempered than in Pibor, however there is still considerable variation and perceptions of safety in Pathai are significantly lower than those in other bomas in the county. Perceptions of safety do not appear to correlate with age group or gender. Qualitative respondents also describe relatively low safety levels in Jonglei. While most of the insecurity is related to cattle raids and, in response, revenge killings, key informants mention the prevalence of both alcohol use and small arms among youth in the region. The combination creates a general feeling of insecurity in moving around the area. JONGLEI Baseline Evaluation | Shejeh Salam Activity 25 “The biggest problem is the number of guns people carry in the cattle camps and when simple quarrel erupts, it leads to conflicts which becomes a big security problem.” Community leader, Jonglei UPPER NILE Most respondents in Upper Nile State indicated that their locality was “Moderately Safe” (47%). At the county level, 32% of respondents in Ulang County indicated that their locality was “Very Unsafe”, compared to only 3% in Baliet County. Respondents in several select Bomas within the Upper Nile state reported high levels of insecurity, with residents reporting “Very Unsafe” conditions in Bomas Waaga (38%), Ying (47%), Toalare (40%), and Kuich (39%) respectively. As with other states, perceptions of safety do not appear to correlate with age group or gender and instead trend most clearly along geographical boundaries. Although perceptions of safety among quantitative respondents are likely driven at least in part by violent expressions of conflict such as cattle raids or revenge killings (see Custom Indicator 3), qualitative respondents remind us that there is a far broader array of considerations. Key informants mention factors such as dangerous roads, human-wildlife conflict, and risk of mosquito-born illnesses. Respondents in Baliet acknowledge that crime – even violent crime – is a factor, but emphasize that it is no more prevalent than in other communities. Like quantitative respondents in Baliet (66% of which said that the county has become more safe over the past year), key informants report improved security conditions. There is, however, an acknowledgement that this peace may be fragile: “There are no safety or security challenges in this community right now, however, we may have safety or security challenges in the future if the other group (Nuer) tries to cause another conflict again.” Community Member (male), Baliet, Upper Nile Baseline Evaluation | Shejeh Salam Activity 26 EASTERN EQUATORIA Perceptions of safety varies considerably at the county level within the Eastern Equatoria cluster. Most respondents (97%) in Budi indicate that their locality was “Very Unsafe” and 94% say that they would not feel comfortable with female family members walking around alone at night. Nonetheless, qualitative respondents suggest that other parts of the county suffer more extreme violence levels, specifically citing bomas in Camp 15 as the most volatile, as communities on the main road are more susceptible to violence. Due to the uniformity of responses among quantitative respondents in Budi, there are no observable trends in terms of gender, age, or ethnicity. In contrast, respondents in the Kapoeta North County indicated that their respective localities were “Moderately Safe” (57%) or “Very Safe” (43%), however there are somewhat differing perspectives on the conditions which create the relatively stable environment in the county. Qualitative respondents associate with the government (at least partially) attributed the relative safety to the presence of formal authorities, such as police and other governing authorities. On the other hand, community members associate the county’s security with strong community cohesion and reliability of the chiefs and/or customary authorities.11 Nonetheless, it is still important to note these perceptions of safety – while vastly better than those of 11 On the flip side, many qualitative respondents attribute fighting and safety issues to drug and alcohol abuse by youth community members. Baseline Evaluation | Shejeh Salam Activity 27 Budi – still reveal some degree of insecurity in the county. More than half (57%) of respondents in Kapoeta North would not feel comfortable with a female family member walking around alone at night in their boma. And, while almost half (49%) of respondent in the county feel that safety in their boma as improved over the past year, a meaningful share (36%) feel that it has become either moderately or significantly less safe over the same period. Disparities in perceptions of safety in the county appear to be determined, to a large extent, by boma of residence. Respondents in Chor and Lole bomas are much more likely to say that their boma is “very safe” than those in Loge or Moru. WESTERN BAHR EL GHAZAL Respondents in WBEG report exceptionally high perceptions of safety, with substantially higher indicator CPI (d) scores than other regional clusters. In Wau county, 87% of respondents describe their boma as “very safe” and most (69%) would feel comfortable with female family members moving around the boma alone at night. Though slightly lower, respondents in Jur River report high safety levels, with 64% describing their bomas as “very safe” and a comparable share (67%) feel comfortable with female family members walking around alone. Indeed, qualitative respondents generally describe their boma as safe, though most cite security concerns in other parts of the region. Some bomas in Wau County, according to a CSO representative, are highly insecure and cautions against traveling there as an outsider, as the “rebels think you are spying on them Baseline Evaluation | Shejeh Salam Activity 28 and they can arrest or kill you.” In these areas, such as Bagari and other places, people must seek approval from the Sudan People’s Liberation Army in Opposition (SPLM-IO) security apparatus. Within counties, there is no discernible difference between male and female respondents’ perceptions of safety, nor does age group appear to predict indicator CPI (d) scores. However, we do find considerable differences between Luo and Dinka respondents in Jur River, whereby perceptions of safety are significantly higher among Dinka respondents than Luo.12 On the other hand, Luo respondents in Wau report higher safety perceptions than Dinka in the county. CPI (E) Definition: Percent of community members reporting having initiated or increased interaction or relationships (in the past 18 months) with members for whom they had strained relationships or past conflict, in Shejeh Salam targeted areas. The relationship could be of any type-trade, cultural (marriage, sport), or even membership of the same peace committee. 12 The difference is statistically significant when controlling for age group and gender. Baseline Evaluation | Shejeh Salam Activity 29 Operationalizing the indicator: In each county the main conflicts – and the primary actors / ethnic groups – were identified (see Annex II). To measure interactions or relationships with previously strained or conflicting groups, we asked respondents about the frequency of marriage, trade, and shared water sources among groups. Using Cronbach’s Alpha test, we find a high correlation (0.79) between the three indicators, suggesting their use as an index is reliable. Table 8: Community Peace Index (e) Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female State Average Southern Unity 0.88 0.84 0.85 0.90 0.91 0.90 0.88 Eastern Equatoria 0.61 0.63 0.62 0.61 0.63 0.62 0.62 Western Bahr el Ghazal 0.77 0.81 0.79 0.80 0.77 0.80 0.79 Jonglei 0.67 0.67 0.67 0.66 0.65 0.65 0.66 Upper Nile 0.80 0.79 0.79 0.76 0.80 0.78 0.79 On average, 75% of respondents report that they have initiated or increased interactions or relationships with members of groups with which they were enemies. Of the various states, Southern Unity has the highest level of interaction with previously hostile groups (88%). In contrast, Upper Niles (79%), WBEG (79%), Jonglei (66%), and Eastern Equatoria (62%) states have moderately lower levels. The Nuer (80%) and Dinka (79%) appear to interact the most with previous conflicting groups compared to other ethnic groups. The Luo and Toposa fall slightly behind at 77% and 66%, respectively. Murle and Buya score comparatively lowest on CPI indicator e, at 62 % and 59 %, respectively. SOUTHERN UNITY Counties in Southern Unity score highly on this indicator compared to those in other clusters. However, this is obscured somewhat by the nature of the conflict in the counties, which exists within subgroups of the Nuer ethnic group. Given this, the extent to which interactions such as trade, marriage, and/or resource sharing exist within conflicting groups is expectedly high – as they are not crossing ethnic groups. Nonetheless, in all three counties, a substantial share of respondents report that shared water resources between Nuer sub-groups is rare. And, although respondents in each county report frequent trade between Nuer sub-groups, intermarriage is relatively rare in Panyijar – with 23% in the county reporting it “never happens.” While age does not appear to influence scores for this indicator, there is a substantial difference between male and female respondents in the cluster.13 Women observe lower levels of inter-group marriage than men but are more likely to report shared resources (water source) than men are. 13 The difference is statistically significant. Baseline Evaluation | Shejeh Salam Activity 30 JONGLEI As a cluster, Jonglei again scores on the lower end for Indicator CPI (e). Scores are lowest in Akobo and highest in Duk and Uror. In Akobo, respondents were asked about the relationship between Lou Nuer, Jikany Nuer, Anywak, and Murle groups. About a third of respondents in the county answered “this never happens” for each of indicator’s subcomponents (when inquired about marriage, trade, and shared water sources among groups), a substantially higher share than in any other county in the cluster. Less than 1% said that the groups frequently trade or share water sources. Indicator scores are comparably low in Pibor, where few respondents say that Murle and Jie ethnic groups frequently marry or share water sources, though more than a quarter (26%) say that intergroup trade is frequent. In Duk and Uror, interactions between previously conflicting groups appears to be much higher, with nearly half of respondents in each of the two counties saying that the groups (Dinka and Nuer in Duk; Lou Nuer in Uror) frequently share water sources. Likewise, more than half say that intergroup trade is frequent. On the other hand, far fewer (36% in Duk and 37% in Uror say that intergroup marriage is common. Although youth and gender do not appear to predict Indicator CPI (e) scores, there is a statistically significant relationship between ethnic group and indicator score: Nuer and Murle score significantly lower than Dinka in Jonglei. EASTERN EQUATORIA Eastern Equatoria performs the poorest of the five regional clusters on indicator CPI (e). In Kapoeta North, participants are responding to questions regarding the Toposa and Boya group relationship. Based on the data, the two groups are unlikely to share water sources or intermarry, with less than 2% of respondents saying either of those is a frequent occurrence. Rather 68% of respondents say that Toposa and Boya never share water sources and 32% say the same about intermarriage between the groups. However, 41% of respondents from the county say that the groups frequently trade with each other. Qualitative respondents largely echo these sentiments, saying peaceful interactions such as these are extremely limited as violence between the Boya and Toposa ethnic groups persists. “For the community to be peaceful, members from different community should be able to work together with those from the other communities, have their children learn together, share the grazing land and no killing anymore. But what we see now here is revenge attack, raiding and more killing of innocent people.” Government Representative, Kapoeta North, Eastern Equatoria In Budi, the Didinga, Boya, and Toposa groups are used to frame inter-group relation questions. Indicator CPI (e) scores are particularly low in Budi, where 99% of respondents say that herders in the two groups never share water sources and 97% said that there is never trade between the groups. Respondents indicate that marriage between the groups is slightly more prevalent, with 50% saying it is rare, but sometimes happens. However, across all three interaction types (marriage, trade, and sharing water sources), less than 2% of respondents in Budi say that it is a frequent occurrence. “For the community to be in peace, we shall gather all the youth from Buya and Toposa and the two communities will be able to stay and Baseline Evaluation | Shejeh Salam Activity 31 graze cattle together not this situation where people are fighting each other to death.” Community Leader, Budi, Eastern Equatoria While there are clear distinctions between the two counties of Eastern Equatoria, neither age group nor gender emerges as a strong determinant of indicator CPI (e) scores in the cluster. WESTERN BAHR EL GHAZAL Respondents in Wau county were asked about conflict between the Luo, Fertit, and Dinka ethnic groups while respondents in Jur River were asked about conflict between Luo and Dinka ethnic groups. Interactions between these groups are relatively frequent, with little variation between counties or interaction types. For each type of interaction (intergroup marriage, intergroup trade, and shared water sources between groups), about two thirds of respondents in both Wau and Jur River report that it is a frequent occurrence. There is no notable difference between male and female or youth and non-youth respondents in WBEG. However, we do find substantial differences between ethnic groups. There is a stark difference between Luo and Dinka respondents in Jur River. Across all three interaction types, Dinka identifying respondents are significantly more likely to say something occurs frequently than Luo identifying respondents. For example, while 85% of Dinka identifying respondents in Jur River say that intergroup marriage between Luo and Dinka occurs frequently, just 48% of Luo identifying respondents say the same. Indeed, 22% of Luo respondents in the county say that intergroup marriage between the groups never happens. This trend is consistent for questions about shared water sources and intergroup trade. Differences between Luo and Dinka identifying respondents in Wau are far less pronounced. Dinka are slightly more likely to say that an interaction occurs frequently than a Luo respondent in the county. UPPER NILE In Baliet, the Dinka Abiliang was the relationship of interest. Respondents indicate a high degree of interactions between these groups, with most (95% or more) respondents reporting that marriage, trade, and sharing water sources are frequent occurrences. In Ulang, participants responded to questions around the Nuer Jikany relationship. In this county, respondents report far lower rates of interaction between the groups. Respondents from Ulang predominantly identify as either Lou Nuer or Nuer Jikany, and there is a sharp contrast between the two groups in relation to indicator CPI (e). Nuer Jikany respondents report significantly higher prevalence of interaction between the two groups – with at least 70% saying each of the interaction types (marriage, trade, shared water source) occurs frequently. On the other hand, just 17% of Lou Nuer respondents say that intergroup trade occurs frequently and in fact 47% say it never happens. These findings make sense when contextualized within the range of conflicts in Baliet County, including its boundary-lands. CUSTOM INDICATOR 2 Definition: Custom Indicator 2 measures the percent of local community members reporting increased trust towards those that they have had a strained or past conflict within the past 18 months. Operationalizing the indicator: In each county, the main conflicts – and the primary actors / ethnic groups – were identified (see Annex 3.) Respondents were asked to comment on the trends in violence and trust between these conflict actors. Custom Indicator 2 measures the share of respondents who state: “the conflict has become less violent than it used to be” and that trust levels between the conflicting groups have increased. Baseline Evaluation | Shejeh Salam Activity 32 The correlation between these two variables is relatively high, suggesting they are measuring related dimensions of intra-community cohesion. However, the indicator is best suited for temporal analysis, rather than geographic analysis. That is, respondents are reflecting on change over time in their county, without referencing to any tangible or absolute levels of distrust that can be compared across locations. Table 9: Custom Indicator 2 Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Total State Average Southern Unity 42% 47% 45% 13% 13% 13% 25% Eastern Equatoria 1% 0% 1% 0% 2% 0% 1% Western Bahr el Ghazal 21% 30% 25% 42% 37% 41% 35% Jonglei 36% 43% 40% 42% 43% 42% 42% Upper Nile 54% 39% 43% 39% 38% 39% 41% The June 2021 Shejeh Salam Baseline survey finds that, overall, relationships between the county’s main conflict groups remain fraught. Across all locations just 29% of respondents report improvements in both violence and tension levels between groups. This is most pronounced in Eastern Equatoria, where nearly 0% of respondents report improvements to violence and tension levels,14 followed by Southern Unity (25%). Of the five regional clusters, respondents in Jonglei are most likely to report improving conditions between conflict groups.15 SOUTHERN UNITY Substantial county-level disparities appear in Southern Unity, where Leer stands out for scoring considerably higher than both Panyijar and Mayendit. In Southern Unity, intra-ethnic conflict exists within the Nuer sub-groups. All respondents in the cluster are Nuer, with Dok-Nuer in Leer, Haak Nuer in Mayendit, and Nyuong-Nuer in Panyijar. As the level of ethnic data available is homogenous at the county level, it is impossible to identify trends within subgroups at a more granular level. However, variation at the boma level suggest differing perceptions of trust among previously conflicting groups. We find significant variation at the boma level in Leer County. In Leer, bomas Dok and Payak score substantially higher on Indicator 2 than bomas Norkong and Padeah. However, in Mayendit and Panyijar, low scores are ubiquitous across sampled bomas. At the cluster level, females score significantly lower on Indicator 2 than male respondents. On the other hand, the difference between youth and non-youth women (and the difference between youth and 14 The difference between Custom Indicator 2 scores in Eastern Equatoria versus the other four regional clusters is statistically significant. 15 The difference between Custom Indicator 2 scores in Jonglei versus the other four regional clusters is statistically significant. Baseline Evaluation | Shejeh Salam Activity 33 non-youth men) is negligible. Overall, non-youth score slightly higher than youth in Southern Unity, but the difference is not statistically significant when controlling for gender. WESTERN BAHR EL GHAZAL Wau and Jur River counties in WBEG have comparable scores for Indicator 2. There is also little variation between age and gender groups, nor are there notable differences between the two counties. Between the two components of the indicator, respondents in both locations are more positive about a drop in violence than improvements in tensions. In Jur River, respondents were asked to reflect upon the Luo-Dinka relationship. While nearly all respondents (88%) in the county confirm a decline in violence, far fewer (35%) report increased trust. It is also important to note the difference between Luo and Dinka identifying respondents in Jur River. For both components of the indicator, respondents who identified as Dinka are substantially more likely to report improvements in the relationship than individuals that identified as Luo.16 This difference is particularly pronounced with regards to changes in trust; nearly half (46%) of Dinka identifying respondents in Jur River report increased trust over the past year, while more than half (59%) of Luo identifying respondents in the county report decreased trust over the same period. The programmatic implications of this finding are such that a reduction in violence is likely a more reliable indicator to measure Shejeh Salam’s initiatives, particularly because levels of trust might still decline even as levels of violence decrease. Moreover, when contextualized within the region’s history, low trust levels might remain. Age and gender appear less reliable predictors of perceptions of trust levels between the Dinka and Luo identifying respondents in Jur River. Youth are more likely to report a decline in trust levels then nonyouth. However, when controlling for ethnic group, the relationship between age group and Indicator 2 is not statistically significant. Similarly, men report lower trust levels than women, but the relationship is again not statistically significant when controlling for ethnic group. In sum, while a respondents age and gender may be related to their perceptions of changes in trust levels between the Luo and Dinka identifying respondents in Jur River, these predictors are far less significant than a respondent’s ethnic group. Ethnic group appears the most important determinant of Indicator 2, whereby Dinka identifying respondents report a more positive relationship than those who identify as Luo. In Wau, respondents were asked about the Luo, Fertit, and Dinka relationships. When asked about trends in violence levels between these groups, most (90%) respondents in the county say it has become violent. Those who say violence has increased are primarily of Krash and Luo ethnic groups. Respondents were much more likely to say that trust has deteriorated between the groups than they were around violence. Nearly a third (30%) said that trust levels have decreased. Again, we see Luo and Krash (and Falata) identifying respondents more likely to report decreased trust levels than other ethnic groups – such as Dinka and Duor. Youth in Wau are more optimistic in their reporting on the conflict and tensions levels than non-youth, as they are more likely to say violence levels are decreasing and trust levels are increasing than older respondents. At the same time, there are significant differences between male and female respondents in the county. Even when controlling for age group, women are significantly more likely to say that trust is increasing between the groups than men are. EASTERN EQUATORIA 16 The difference is statistically significant. Baseline Evaluation | Shejeh Salam Activity 34 Indicator 2 scores lowest in Eastern Equatoria, where few respondents report an increase in trust levels. This is consistent in both counties in the cluster (Kapoeta North and Budi) and across demographic groups. It is far more pronounced, however, in Budi, where most respondents report that tension and violence between Didinga, Boya, and Toposa groups have worsened. In fact, 81% of respondents in Budi say that theirs has become more violent over the past year and 85% report decreased trust among groups. Though this is somewhat more tempered in Kapoeta north, about half (49%) of respondents in the county report decreased trust between groups and 50% say that it has become more violent. It is also important to note differences between demographic groups in Eastern Equatoria. Women in the cluster are more likely to report a decrease in violence than men are17 while youth are more likely to report a decline in trust between groups than non-youth.18 As ethnic groups are homogenous within counties, it is difficult to ascertain the extent to which geography and/or ethnic identity influence perceptions of trust between strained groups. UPPER NILE While Upper Nile is amongst the highest on this indicator, there is considerable variation between Baliet – whose score appears to be driving the high indicator score at the cluster level - and Ulang – where trust and violence levels appear to be in decline. In Baliet, respondents identify as Padang Dinka and are asked to respond to questions around the Dinka Abiliang conflict. Most (86%) say that the conflict has become less violent and that trust levels have increased (91%). Interestingly, although the overall share of respondents reporting increased violence in Baliet is low, women in the county are significantly more likely to report increased violence levels than men.19 We find no real difference between youth and non-youth respondents in the county. In Ulang, the Nuer Jikany conflict is of interest and more than half (56%) in the county say that the relationship has become more violent over the past year. A comparable share (51%) report a deterioration in trust levels over the same period. While we find no difference in reported violence levels between male and female respondents in the county, male respondents are more likely to report decreased trust than female respondents.20 At the same time youth – regardless of gender – are report worsening trust levels more frequently than non-youth in Ulang do.21 Among the two main ethnic groups in Ulang – Lou Nuer and Nuer Jikany – there are also substantial differences in perceptions of trust and violence between the groups. Nuer Jikany are more likely to say that violence has increased and trust levels have worsened than Lou Nuer identifying respondents.22 JONGLEI Among the 5 clusters, Jonglei performs higher on Indicator 2. At the cluster levels, neither age group nor gender are significant predicators of Indicator 2 scores. There is meaningful variation among counties: Uror county is unique within the cluster, with a relatively low score compared to the other 17 The difference is statistically significant when controlling for age group and county. 18 The difference is statistically significant when controlling for gender and county. 19 The difference is statistically significant when controlling for age group. 20 The difference is statistically significant when controlling for age group. 21 The difference is statistically significant. 22 The differences are statistically significant when controlling for gender and age group. Baseline Evaluation | Shejeh Salam Activity 35 counties within the cluster. However, it is difficult to parse trends beyond that without further qualitative data. That is, perceptions of increased trust among recently conflicting groups does not appear to vary substantially by ethnic group, boma, gender, or age. CUSTOM INDICATOR 3 Definition: Percent of community members reporting that key conflicts (namely cattle raid, revenge killing, child abduction, or rape) have reduced in the past 18 months. Operationalizing the indicator: Respondents were asked whether they know anyone in the past year who has been affected by: cattle raids, revenge killings, rape, and child abductions. The indicator is calculated based on the mean of these four indicators – or as a representation of the number of times a respondent said “yes” they know someone who was affected by the violence type. The score is between 0 and 1, where a score of 0 indicates a respondent said “no” they do not know anyone who has been affected by that, to all four violence types. A score of 1 indicates a respondent said “yes” to all four violence types. For each violence type that a respondent said “yes” they know someone who has been affected by it, we followed up by asking whether the incident occurred in their boma. This is done as a method of better gauging frequency of such events –as many people may report awareness of a few high-profile events at the county level. In this case, many people reporting on a few high-profile events may be misinterpreted as a high frequency. Table 10: Custom Indicator 3 Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Total State Average Southern Unity 0.52 0.45 0.48 0.56 0.47 0.53 0.51 Eastern Equatoria 0.52 0.56 0.54 0.61 0.52 0.58 0.56 Western Bahr el Ghazal 0.19 0.21 0.20 0.17 0.15 0.16 0.18 Jonglei 0.64 0.60 0.61 0.56 0.55 0.56 0.58 Upper Nile 0.40 0.44 0.43 0.41 0.42 0.42 0.42 Overall, most (75%) of respondents reportedly faced at least one of the four key conflicts, that is cattle raid/theft, revenge killing, child abduction and or rape. Amongst the five states, Eastern Equatoria , Jonglei and Southern Unityare the most affected states. The average score for Upper Nile falls only slightly below and Western Bahr-el Ghazal seems comparatively less affected by any of the conflict types. The Buya, Murle, Nuer and Toposa groups rank above average; these groups seem to be more affected by key conflicts than are the Dinka and Luo. Baseline Evaluation | Shejeh Salam Activity 36 SOUTHERN UNITY Respondents in Southern Unity suggest relatively high levels of violence in the region. Most (88%) respondents know someone who has been affected by at least one of the violence types listed (cattle raids, revenge killings, rape, and child abductions). This is highest in Mayendit, where less than 1% of respondents say they are not aware of anyone who was affected by any of the violence types in the past year. In fact, as many as 18% of respondents in Mayendit answered “yes” to all four violence types; in other words, nearly a fifth of people in Mayendit report knowing at least one victim of each of the listed violence types over the past year. This is considerably higher than Leer and Panyijar, though even in these counties, most respondents know victims of two or more of the violence types. It is also important to note that many of the instances occurred in the respondent’s boma of residence. Nearly all (99%) of the violent incidences referenced by respondents in Mayendit occurred in their boma. Again, while this is somewhat lower in the other two counties, the share of respondents who say that these violent incidences occurred in their county remains high: at 89% in Panyijar and 77% in Leer. Women in Southern Unity report awareness of violent incidences at a higher rate than men. Even when controlling for county, women are more likely to say they know someone who has been affected by violence than men.23 This difference is starkest with regards to revenge killings. There is a 13-point difference between men and women on this variable, with 69% of men and 82% of women in Southern Unity saying that they know someone who has been affected by revenge killing over the past year. There is also a statistically significant difference between youth and non-youth scores on Indicator 3, whereby youth are more likely to report awareness of violent instances than non-youth.24 Although the difference between male and female scores for Indicator 3 in Southern Unity is driven primarily by the question about revenge killings, significant differences between youth and non-youth responses are observed across all violence types, apart from child abductions. Of the violence types, cattle raids and revenge kills appear most common in Southern Unity, with 82% and 78% of respondents in the region saying they know someone who has been affected, respectively. This is supported by qualitative data, in which most respondents are aware of one or more recent cattle raids and some describe it as the most prominent security-related issue in the region. On the other hand, just 10% of respondents know someone who has been affected by child abduction (though this figure is as high as 20% in Mayendit). Likewise, most key informants describe child abduction as a rare occurrence in Southern Unity. However, at least one key informant knew of a child abduction incident occurring this year. Just as child abductions rates fall lower than cattle raids and revenge killings, so do rape incidence rates. However, 34% of Southern Unity respondents know someone being affected by rape. The figure stands highest in Panyijar, at 44%, compared to Mayendit and Leer Counties. Although Mayendit stands out within Southern Unity for higher reported incidences of violence, respondents’ perceptions of trends – i.e. the increasing or decreasing prevalence of such events – complicates the picture somewhat. Respondents in Mayendit are – for all violence types – more likely than those in Leer and Panyijar to say that the event is becoming less frequent. This is particularly relevant with regards to cattle raids; while 63% of respondents in Mayendit report a decline in cattle raids over the past year, a similar share (61%) in Panyijar report an increase over the same period. A similar trend emerges with regards to revenge killings, whereby 68% in Mayendit say that revenge killings are occurring less frequently while 66% in Panyijar say the opposite. 23 The difference is statistically significant. 24 The difference is statistically significant, even when controlling for gender and county. Baseline Evaluation | Shejeh Salam Activity 37 JONGLEI Jonglei has the highest score among the five regional clusters for Indicator 3, meaning it is the cluster with the highest level of reported violence. Nearly all (87%) respondents say that they know of at least one person who has been affected by one or more of the violence types. Within the cluster, this is lowest in Akobo County, where more than a quarter (26%) of respondents say that they do not know anyone who has been affected by any of the violence types inquired about. It is highest in Duk, where nearly half (48%) know at least one person who has been affected by all four violence types. Most of the violence people are aware of happened in their boma of residence, with 82% of violent instances mentioned confirmed by respondents to have occurred within their boma. We also find that men report significantly higher levels of violence than women on questions related to Indicator 3. Even when controlling for county variation in Jonglei, the difference between male and female scores on Indicator 3 is statistically significant. Although women report a lower awareness of violence across all four violence types, the difference is starkest with regards to revenge killings and cattle raiding. Different age groups – youth and non-youth – do not appear to have any meaningful differences in Indicator 3 scores in the cluster. “This is the number one concern here, because cattle means life to so many people here.” Community Leader, Jonglei “Cattle raiding is the most common security and safety threat in the county” Community Member, male, Uror, Jonglei Cattle raids appear to be the most prominent of the four violence types in Jonglei; for instance, 73% of respondents in the cluster know someone who has been affected by cattle theft – 87% of whom confirm it has occurred in their boma of residence. The share of respondents who know someone who has been affected by child abductions is also high (65%). A similar share (62%) knows someone who has been affected by revenge killings in the past year. Again, nearly all respondents in Jonglei who know someone who has been affected by this violence type confirm that it occurred in their boma (82% for child abductions and 85% for revenge killings). Rape cases are again reported at lower rates than other conflict types but still make up for a remarkable 30%. Across all violence types, Duk reports the highest levels of violence – with 98% of respondents knowing someone who has been affected by cattle raids. EASTERN EQUATORIA Respondents in Eastern Equatoria report high levels of violence. Most qualitative respondents in Eastern Equatoria could recall an incident of cattle theft within the past few months, several within the weeks preceding data collection. Indeed, cattle raiding and revenge killings between Buya and Toposa ethnic groups are relatively common. Baseline Evaluation | Shejeh Salam Activity 38 Quantitative survey participants confirm these findings. Of the four types of violence presented to participants, nearly all (98%) in Budi said that they were aware at least two of these occurring in their county over the past year. This is driven primarily by two violence types – cattle theft and revenge killings – for which almost all (98% and 99% respectively) respondents know someone who has been affected by in the past year. It is also worth noting that most respondents (70%) say that the frequency of revenge killings and cattle raids has increased in Budi over the past year. And even though rape and child abduction appear far less common in the county, just as they are in Southern Unity and Jonglei, they still account for a quarter of respondents. In other words, 24% of respondents know someone who has been affected by rape in the past year and 23% for child abductions. While the question asks about incidences in the county, respondents who confirm “yes” they know someone who has been affected by the violence type are asked a follow up question as to whether the incident they are thinking of occurred in their boma of residence. In Budi, 98% confirmed that the incident occurred in their boma. These figures are somewhat lower in Kapoeta North, though still reveal a high prevalence of violence: 81% in the county are aware of at least one of the violence types occurring in their county over the same period. Cattle raids and revenge killings are also prevalent in Kapoeta North, according to respondents, 73% in the county know someone who has been affected by cattle raids in the last year and 65% for revenge killings. Like in Budi, we find that respondents are less aware of instances of rape and child abductions in the county, with 36% and 33%, respectively, of respondents reporting awareness of such events. Perhaps the most substantial difference between Budi and Kapoeta north is the share of incidences that occurred in the respondent’s boma of residence. While nearly all respondents (98%) reported the violent incidences in Budi occurred in their boma, just 28% of those in Kapoeta North. Although this is an imperfect indicator, it does suggest the prevalence of violence in Kapoeta North is even lower than reported, relative to Budi. WESTERN BAHR EL GHAZAL Indicator 3 scores are particularly low in WBEG, translating into relatively low levels of violence in the cluster – according to respondents. More than half of respondents (62% in Wau and 56% in Jur River) say that they have not heard of anyone in their county affected by any of the violence types in the past year. Of the violence types, child abductions appear the least prevalent – with just 5% in the cluster reporting a case occurring in the past year. The remaining violence types – revenge killings, cattle raids, and rape – are relatively comparable, with 20-25% of respondents in the cluster reporting awareness of a case in the past year. In other words, 19% of respondents are aware of revenge killings, 21% report a rape case and 25% know of cattle raids. Additionally, respondents overwhelmingly report a downward trend of violence, with all types of violence occurring less frequently than the prior year. Western Bahr el Ghazal is similar to other regional clusters, in that women report lower levels of violence then men in the same county.25 And, like other clusters, there is no meaningful difference between youth and nonyouth reporting of violence on questions relating to Indicator 3. However, WBEG does stand out in the consistency with which ethnic group identities correlate with indicator scores. Dinka in Jur River report lower levels of violence than Luo.26 In Wau, potential differences between ethnic groups are less clear. At the same time, qualitative respondents classify cattle raids as a high priority in the area and several listed incidences which occurred in WBEG within the past 6 months. 25 The difference between male and female scores for Indicator 9 are statistically significant when controlling for age and county. 26 The difference is statistically significant when controlling for gender and age. Baseline Evaluation | Shejeh Salam Activity 39 “Cattle raiding is very big issue in Jur River County and people are very concerned about it. Last time cattle raid happened in February this year between the Lous and Dinka from Warrap.” Government Representative, Jur River, Western Bahr el Ghazal UPPER NILE There is considerable variation in Indicator 3 scores between counties in Upper Nile, with respondents in Ulang reporting substantially higher levels of violence than those in Baliet. While more than half (53%) of respondents in Baliet say they do not know anyone affected by any of the violence types in the past year, just 14% in Ulang say the same. In fact, it is more common, in Ulang, to know at least one person who has been affected by all four violence types (28%). Reported instances are highest among revenge killings (79%) and cattle thefts (74%) in Ulang, though prevalence rates of rape and child abductions are still relatively high (41% and 49%, respectively). The share of respondents in Baliet who know someone who has been affected by revenge killings is much lower – at 21%. Though respondents are somewhat more likely to know someone affected by cattle raids in Baliet (37%) these figures are lower for rape (12%) and child abductions (16%). Though qualitative respondents in Upper Nile are all aware of incidences of cattle raids, most say that it occurs in neighboring communities – or is driven by outsiders who come into their communities and engage in livestock theft. Several respondents refer to an event in January 2021 in which several hundred cattle were stolen and which resulted in the death of multiple people – according to key informants. And, although several key informants cite incidences where cattle raids did not result in violence against humans, most (69%) quantitative respondents in Upper Nile say that, following a cattle raid, people are likely to be killed for revenge. “…it occurred in Adong payam of Baliet county where cattle were stolen by Nuer Lou in May 10, 2021 and there was not violence because the cattle thieves left the stolen cattle in the bush and the owners came after them and took their cattle back. No one retaliated after that!” CSO Representative, Baliet, Upper Nile CUSTOM INDICATOR 9 Definition: The percent of local community members who believe that traditional authorities are playing their role in conflict prevention, resolution, or mitigation Operationalizing the indicator: Respondents were asked whether they think that traditional authorities are effective in maintaining peace in the community. Baseline Evaluation | Shejeh Salam Activity 40 Overall, more than three-quarters (78%) of the population in the Shejeh Salam intervention areas believe that traditional authorities are effective in keeping peace. This view is strongest in Southern Unity, at 94%. Among the ethnic communities, the Toposa and Dinka in particular find that traditional authorities deal effectively with peace (98 % and 92 %, respectively). The Nuer (76 %), Murle (74 %) and Luos (67 %) expose slightly lower levels of trust. Comparatively fewer Buya believe that peace work of traditional authorities is effective (55 %). Table 11: Custom Indicator 9 Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Total State Average Southern Unity 92% 93% 92% 96% 92% 94% 94% Eastern Equatoria 74% 78% 76% 77% 79% 78% 77% Western Bahr el Ghazal 68% 66% 67% 79% 73% 77% 74% Jonglei 81% 85% 83% 68% 57% 63% 70% Upper Nile 87% 80% 82% 71% 76% 74% 77% Indicator 9: Effectiveness of Customary Authorities SOUTHERN UNITY Most (94%) respondents feel traditional authorities are effective. There is little variation between counties on this, though respondents in Mayendit are somewhat less likely to say that traditional authorities are effective in maintaining peace. This may be owed, in part, to the fact that the county generally experiences a greater degree of conflict and violence (based on data from Indicator 3) than Leer and Panyijar. In this sense, there is a general failure to maintain peace in the county. However, it is important to note that customary authorities are still the preferred actor that community members in Mayendit turn to for conflict prevention and/or resolution. Neither gender nor age group appear to influence Indicator 9 scores. 0% 2% 3% 6% 7% 6% 21% 23% 25% 16% 94% 77% 74% 70% 77% Baseline Evaluation | Shejeh Salam Activity 41 EASTERN EQUATORIA Counties within Eastern Equatoria have starkly divergent perceptions of the effectiveness of traditional leaders in maintaining peace. In Kapoeta North, nearly all (98%) respondents say that they are effective and 99% want traditional elders to be more involved. Alternatively, just 55% of respondents in Budi feel that they are effective. While opinions are not necessarily negative (just 3% in Budi said that traditional leaders make things worse), a large share (41%) feel that their involvement has no effect at all on the maintenance of peace in the community. Qualitative respondents generally agree that the ceremonial nature of the traditional leaders has implications on their effectiveness in resolving violent conflicts, as the extent to which they can enforce decisions and/or consequences is limited. They are unable to imprison offenders, as one respondent explains as an example of their enforcement deficits. Nonetheless, there are measurable benefits to resolving conflicts through traditional authorities over police, as respondents claim chiefs are quick to attend to the issue, their priority is on maintaining good relations between groups, and the police – unlike chiefs – may charge constituents for justice-related services. “The chiefs are incapable of dealing with violent and disobedient individuals because we don’t have the capacity to enforce the law, that is why we are requesting the police to be deployed here in order to deal with difficult issues.” Community Leader, Budi, Eastern Equatoria WESTERN BAHR EL GHAZAL Most (74%) respondents in WBEG say that traditional leaders are effective in maintaining peace. This figure is somewhat higher in Wau (79%) than it is in Jur River (68%). The vast majority of respondents (92% in Jur River and 96% in Wau) want traditional leaders to be more involved in community peacebuilding in the future. Qualitative respondents in WBEG echo many of the same sentiments as those in other regional clusters, describing customary authorities as efficient and fair, preferred by many because they are embedded within the community. However, unlike other regions, at least one key informant in Jur River claims that customary authorities may charge substantial fees to resolve a dispute. In other clusters, key informants generally said that customary authorities were less expensive than other conflict mitigation actors, such as the police. Although perceptions of the effectiveness of traditional leaders is relatively high for all demographic groups, we do find some differences between male and female respondents in WBEG. Neither male nor female respondents are likely to say that traditional leaders make things worse (just 3% of respondents said as much). However, men are more likely than women to say that traditional leaders have no effect.27 We also find differences between ethnic groups in the state, whereby respondents identifying as Luo report lower levels of effectiveness among traditional leaders.28 JONGLEI Perceptions of the effectiveness of traditional leaders vary considerably across counties. They appear weakest in Akobo; 22% of respondents in the county say that the involvement of customary authorities 27 The difference is statistically significant when controlling for age group and county. 28 The difference is statistically significant when controlling for gender and county. Baseline Evaluation | Shejeh Salam Activity 42 makes things worse and another 56% say they have no effect at all. In Duk and Uror counties, on the other hand, the majority (95% and 89%, respectively) feel that customary authorities are effective in the maintenance of peace. Perceptions in Pibor are somewhat more mixed; although most (74%) feel customary authorities are effective, the remaining 26% say that they have no effect. Among respondents who feel that customary authorities are effective, 90% feel that they should be more involved than they currently are. Although customary authorities are not perceived to be particularly effective in Akobo County, they remain the most commonly cited actor that respondents would turn to for support in preventing conflict escalation. After the county commissioner, they are also the most popular actor respondents would turn to for the resolution of violent conflict. Customary authorities are also the primary mechanism for conflict resolution and the prevention of violence in Pibor and Uror. In Duk, most respondents would turn to the police instead. Female respondents in Jonglei perceive traditional leaders as less effective than men.29 The indicator does not vary between age groups in Jonglei. UPPER NILE The Upper Nile cluster includes respondents from Baliet and Ulang counties. In Baliet, 95% of respondents say that traditional leaders are effective in maintaining peace in the community and 85% would like to see them become more involved than they currently are. Respondents in Ulang, while generally positive, are less uniform in their perceptions of the effectiveness of traditional leaders. While 61% in the county feel that they are effective, 27% say traditional leaders have no effect and 12% say that their involvement makes things worse. This is driven primarily by Lou Nuer respondents, who (among the predominant ethnic groups in the county) are generally more skeptical of the utility of traditional leaders than Nuer Jika respondents in Ulang.30 Both gender and age group also appear to relate to perceptions of traditional elders as effective peacekeepers. Women are more likely to say traditional elders are effective than men are and non-youth more likely to say they are effective than youth in Upper Nile.31 CUSTOM INDICATOR 19 Definition: Percent of community members with adequate knowledge of trauma in the community. Operationalizing the indicator: Indicator 19 is comprised of three sub-components. Cumulatively, the Indicator measures the prevailing state of trauma awareness within the community, including a sense of penetration of trauma-related interventions by Shejeh Salam and other actors in Shejeh Salam targeted areas. The three dimensions, or sub-components, of the indicator are as follows: a. Percent of community members with adequate knowledge of trauma – including both causes and symptoms or effects of trauma. b. Of the community members with adequate knowledge of trauma, the percent who can relate trauma with aggressive behavior and conflict. 29 The difference is significant, even when controlling for county. 30 The difference is statistically significant when controlling for gender and age group. 31 The differences are statistically significant. Baseline Evaluation | Shejeh Salam Activity 43 c. Of the community members with adequate knowledge of trauma, the percent reporting that there are some trauma related interventions in the community For Indicator 19 (a), adequate knowledge of trauma is determined through a two-stage evaluation. In the first state, respondents were presented a list of 11 items and asked to determine whether the item is a cause of trauma. Respondents who correctly classified (as either TRUE, a cause of trauma or FALSE, not a cause of trauma) at least six items (more than 50%) were determined to have adequate knowledge of trauma causes. In the second stage, a similar exercise presented respondents with a list of 10 items and asked them to determine whether the item may be an effect/symptom of trauma. Respondents who correctly classified (as either TRUE, an effect of trauma or FALSE, not an effect of trauma) at least six items (more than 50%) were determined to have adequate knowledge of trauma effects/symptoms. These two stages of the evaluation were then used to produce a score for Indicator 19 (a). If a respondent answered at least 50% correct in both (cause and effect) stages, they are deemed to have an adequate knowledge of trauma. Table 12: Indicator 19a Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Average Southern Unity 1.00 0.99 0.99 0.99 1.00 1.00 0.99 Eastern Equatoria 1.00 1.00 1.00 1.00 0.98 1.00 1.00 21% 74% 84% 0% 20% 40% 60% 80% 100% Knowledge of interventions in the community (19c) Abiity to relate trauma with conflict / aggression (19b) Knowledge of trauma causes, effects, and symptoms (19a) Custom Indicator 19 Baseline Evaluation | Shejeh Salam Activity 44 Western Bahr el Ghazal 1.00 0.95 0.98 0.96 0.96 0.96 0.96 Jonglei 0.62 0.51 0.55 0.55 0.71 0.62 0.60 Upper Nile 0.50 0.37 0.41 0.52 0.58 0.55 0.49 While the majority of 80% of respondents in Shejeh Salam targeted areas have adequate knowledge of trauma, there are regional disparities. Indicator 19 (a) scores are extremely high in Southern Unity, Eastern Equatoria, and WBEG, with nearly all respondents demonstrating an adequate knowledge of trauma. Scores are lower in Jonglei and Upper Nile, where there is considerable variation at the county level. In Jonglei, for example, nearly all (98%) respondents in Akobo demonstrate adequate knowledge of trauma while the same is true of just 4% in Duk. Pibor and Uror fall somewhere between these two poles, at 55% and 85%, respectively. Although there is no discernible difference between knowledge of trauma and age, we do find women perform better than men on Indicator 19 (a).32 Differences also appear between ethnic groups. In each ethnic group, more than half of the respondents have trauma knowledge. However, respondents from the Toposa, Buya, and Luo groups score the highest with 100%, 99.5%, and 92.3%, respectively. Nuer, Dinka and Murle have comparatively lower trauma knowledge levels at 80%, 62% and 55% respectively. For example, while the Nuer in Southern Unity, Eastern Equatoria, and Jonglei score the highest with at least 90% in these states, Nuer respondents in Upper Nile have comparatively low trauma knowledge with only 35.6%. Notably, elderly females score highest among the Nuer in Upper Nile (48.4%), non-adolescent male respondents have the lowest trauma knowledge (22.2%). A similar trend is seen among the Dinka. The survey found that they have adequate trauma knowledge in Southern Unity, Eastern Equatoria, and WBEG states (at least 98%), but lower trauma knowledge in Jonglei and Upper Nile (1.4% and 65.1%, respectively). Amongst Dinka in Upper Nile, women perform better than men, especially young women (74.6%) have the best knowledge, while young men (55.2%) know the least. Indicator 19 (b) pertains only to those with an adequate knowledge of trauma and asks respondents to tell us to what extent they agree with the following statement: “Trauma is a significant cause of violence and conflict in communities.” The indicator, which measures ability to relate trauma with aggressive behavior and conflict, counts the share of respondents who answer either “to some extent” or “very much so” to the above question. Table 13: Indicator 19b Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Total State Average Southern Unity 0.96 0.99 0.98 0.98 1.00 0.99 0.98 Eastern Equatoria 1.00 1.00 1.00 1.00 0.98 1.00 1.00 32 The difference is statistically significant when controlling for regional cluster and age. Baseline Evaluation | Shejeh Salam Activity 45 Western Bahr el Ghazal 0.95 0.91 0.93 0.87 0.80 0.85 0.88 Jonglei 0.45 0.38 0.40 0.34 0.40 0.36 0.38 Upper Nile 0.48 0.36 0.39 0.49 0.53 0.51 0.46 As may be expected, a similar trend emerges for Indicator 19 (b), whereby Jonglei and Upper Nile lag behind the other regional clusters in their understanding of trauma and its relationship to violence and conflict. However, in this case neither gender nor age are significant predictors of Indicator scores. Different ethnic groups associate trauma with conflict violence to varying degrees. Almost all respondents from the Buya and Toposa groups are able to associate trauma with aggressive behavior (99.5% and 100%, respectively). Fewer respondents from the Luo(78.3%) and Nuer (68.2%) are able to do so, and the Dinka and Murle are the least able to associate trauma and violence (58.7% and 55%, respectively). The overall trend for Nuer and Dinka reverses in Jonglei and Upper Nile states. In both states, only 46% and 32.7% of Nuer, and only 1.4% and 14.3%, respectively, of Dinka acknowledge trauma as a cause of violence. Indicator 19 (c) also pertains only to those respondents found to have an adequate knowledge of trauma and asks whether there are any people, groups, or organizations in the community who work to help reduce the symptoms or impact of trauma. The measures the share of respondents who report that there are some trauma-related interventions in the community. All states scored relatively similarly on indicator (c). On average, 21% of respondents per state indicated that they knew of a project to reduce the symptoms or effects of trauma. Differences exist across ethnic groups. Murle and Buya are well below and Toposa above the average (5.5%, 3.9%, and 49.4%, respectively). The overall trend for Dinka (17.9%) and Nuer (20.9%) varies slightly across states. Most notably, in Eastern Equatoria, WBEG, Jonglei, and Upper Nile states, none of the Dinka respondents know of any measures. Among the Nuer, fewer than 7.9% of respondents in Upper Nile can name interventions, falling below the state average. Table 14: Indicator 10c Scores Regional Cluster Male Youth Male Non￾youth Male Total Female Youth Female Non￾youth Female Total State Average Southern Unity 0.15 0.21 0.19 0.23 0.20 0.22 0.21 Eastern Equatoria 0.20 0.29 0.25 0.24 0.35 0.28 0.26 Western Bahr el Ghazal 0.23 0.27 0.25 0.19 0.20 0.19 0.21 Jonglei 0.21 0.11 0.15 0.19 0.27 0.22 0.20 Upper Nile 0.24 0.12 0.16 0.15 0.14 0.15 0.15 Given components b and c are filtered to only include respondents of a sufficient score on component a, the scores are inherently related and measuring their covariance is not a helpful measure of the indicator’s reliability. Baseline Evaluation | Shejeh Salam Activity 46 CONTEXT INDICATORS The Shejeh Salam Context Indicators seek to understand the nature of media access and trust among community members in target counties. Access to news and media is driven, in large part, by geography. This is unsurprising, given the variation in infrastructure reach and quality between locations. Respondents in Wau, Jur River, Duk Baliet, and Ulang counties are more likely to be able to easily access a radio than respondents in other counties.33 On the other hand, respondents in Budi are significantly less likely to answer yes – with just 1% of respondents reporting to have easy access to a radio any time they want to listen to a program. Only three counties (or 23%) of the 13 counties have radio access of 60% or higher with Pibor having the highest at 80%, followed by Ulang at 79% and Wau at 66%. Six of the 13 counties (nearly half or 46%) have radio access of 20% or lower with Akobo and Budi having less than 10% access. The remaining four counties (nearly a third or 31%) have radio access ranging from 20% to 48%. In Wau, the relatively high radio access (66%) is associated with very high ranking of radio as the number 1 trusted source of information. Still in WBEG, Jur River has a 31% radio access and radio is also ranked as the number 1 trusted source. Although radio access in Pibor is the highest at 80%, and in Uror it is at 40% which is also much higher than in many other counties, radio is ranked almost last as the most trusted source. In Ulang, despite high radio access is at 79%, radio is still ranked averagely (number 4 like) just like in counties with low radio access. However, geography is not the sole driver behind radio access. Even when controlling for regional cluster, both gender and age group have statistically significant relationships to access to radio, where women are less likely to have access than men and youth are more likely to have easy access to a radio than non-youth. While 32% of male respondents say that they have regular access to a radio, just 14% of female respondents say the same. The gap between youth and non-youth is smaller (just 1 percentage point different) but still significant. Other socioeconomic characteristics also appear to be meaningful predictors of regular access to a radio, though these are expected trends. Respondents who identify as businessmen/women and those with waged or salaried employment are more likely to have regular access to a radio than those who identify as farmers, herders, or unemployed. Likewise, respondents who are married are more likely to have access to a radio than those who are singly or widowed. Such trends generally confirm that radio access tends to fall along predictable socio-economic lines. Where radio programming intends to target specific groups – for example cattle herders or farmers – the constraints of radio access should be taken into consideration as a potential challenge or limitation. Table 15: Context Indicators, media access and exposure % "yes" to: "Can you easily access a radio any time you want to listen to a program?" % who listen to the radio regularly (more than 3 days per week) % who listen to the news regularly (more than 3 days per week) Southern Unity 13% 7% 8% Mayendit 10% 4% 5% Leer 18% 10% 12% 33 The difference is statistically significant. Baseline Evaluation | Shejeh Salam Activity 47 Panyijar 12% 6% 7% Eastern Equatoria 7% 3% 3% Kapoeta North 12% 5% 5% Budi 1% 0% 0% Western Bahr el Ghazal 48% 35% 37% Wau 66% 17% 53% Jur River 31% 53% 20% Jonglei 11% 6% 5% Akobo 7% 1% 1% Duk 24% 14% 3% Pibor 80% 4% 7% Uror 40% 2% 8% Upper Nile 24% 6% 12% Baliet 30% 13% 23% Ulang 79% 1% 2% Trust in radio as a news source is, unsurprisingly, closely linked to listenership. Respondents who listen to the radio at higher rates are more likely to select radio as their most trusted new source. However, in many locations, respondents indicate higher trust levels in individuals they have relationships to, such as relatives, religious leaders, and/or traditional leaders. This suggests a need for messaging campaigns to incorporate key influential figures at the community level, in addition to radio programming. It is important to note, however, the ubiquitous distrust of social media and political leaders as reliable purveyors of news and/or information. Annex V presents the ranking of news sources based on the percent of respondents reporting that it is the most trusted source. Baseline Evaluation | Shejeh Salam Activity 48 Table 16: Context Indicators, trusted sources Radio Relatives Friends Religious Leaders Political Leaders Traditional Leaders Social Media Southern Unity 9% 33% 15% 23% 1% 17% 1% Mayendit 7% 50% 13% 16% 2% 12% 1% Leer 15% 23% 22% 7% 2% 29% 3% Panyijar 6% 28% 8% 47% 0% 8% 0% Eastern Equatoria 3% 31% 13% 1% 1% 27% 0% Kapoeta North 6% 30% 13% 2% 1% 50% 0% Budi 1% 32% 14% 0% 1% 3% 0% Western Bahr el Ghazal 46% 18% 6% 18% 0% 8% 2% Wau 63% 9% 8% 11% 0% 4% 3% Jur River 28% 27% 5% 24% 1% 11% 1% Jonglei 4% 37% 5% 19% 5% 28% 0% Akobo 2% 36% 1% 21% 0% 39% 1% Duk 12% 14% 11% 36% 4% 17% 0% 0% 20% 40% 60% 80% 100% Relatives Friends Traditional Leaders Religious Leaders Radio Political Leaders Social media - facebook Newspaper News Sources - Use and Trust Use or receive news from Most trusted new source Baseline Evaluation | Shejeh Salam Activity 49 Pibor 0% 48% 3% 3% 12% 33% 0% Uror 1% 52% 5% 16% 1% 21% 0% Upper Nile 15% 37% 6% 19% 1% 19% 1% Baliet 20% 44% 5% 9% 1% 19% 1% Ulang 10% 31% 6% 28% 2% 19% 1% Baseline Evaluation | Shejeh Salam Activity 50 ANNEX 1: INDICATOR PERFORMANCE TABLE Indicator CPI (a) Indicator: Percent of community members who answer “yes” to the following question: “As far as you know, over the past year have there been any new peacebuilding systems or projects in your community? These can include people or groups who try and prevent conflict, resolve disputes, or maintain peace in the community. They can include initiatives by people in the community .” FACTOR Estimate P-value Significance Pairwise analysis Overall 0.256 Southern Unity 0.160 0.135 Southern Unity compared to all other regional clusters Eastern Eq. 0.093 0.001 ** Eastern Eq. compared to all other regional clusters WBEG 0.347 0.070 * WBEG compared to all other regional clusters Jonglei 0.251 0.925 Jonglei compared to all other regional clusters Upper Nile 0.417 0.011 ** Upper Nile compared to all other regional clusters Dinka 0.300 0.445 Dinka compared to all other ethnic groups Nuer 0.313 0.190 Nuer compared to all other ethnic groups Luo 0.395 0.102 Luo compared to all other ethnic groups Murle 0.033 0.000 ** Murle compared to all other ethnic groups Buya 0.022 0.000 ** Buya compared to all other ethnic groups Toposa 0.167 0.103 Toposa compared to all other ethnic groups Other 0.248 0.881 Other compared to all other ethnic groups Male_Youth 0.227 0.216 Male youth compared to all other demographic groups Male_Older 0.232 0.398 Male non-youth compared to all other demographic groups Female_Youth 0.248 0.592 Female youth compared to all other demographic groups Female_Older 0.312 0.022 ** Female non-youth compared to all other demographic groups Southern Unity 0.160 Dok Nuer 0.023 0.049 ** Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.000 0.163 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.058 0.172 Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator CPI (a) Indicator: Percent of community members who answer “yes” to the following question: “As far as you know, over the past year have there been any new peacebuilding systems or projects in your community? These can include people or groups who try and prevent conflict, resolve disputes, or maintain peace in the community. They can include initiatives by people in the community .” Baseline Evaluation | Shejeh Salam Activity 51 Female Youth 0.000 0.148 Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Older 0.000 0.147 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.136 0.820 Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.188 0.289 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Male Older 0.059 0.468 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.128 0.248 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Older 0.158 0.555 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 0.331 0.005 ** Nyuong Nuer in Southern Unity compared Unity compared to other ethnic groups in Southern Unity Male Youth 0.125 0.097 * Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 0.455 0.454 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 0.323 0.838 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 0.353 0.556 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.093 Buya 0.017 0.047 ** Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.000 0.387 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.000 0.332 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Indicator CPI (a) Indicator: Percent of community members who answer “yes” to the following question: “As far as you know, over the past year have there been any new peacebuilding systems or projects in your community? These can include people or groups who try and prevent conflict, resolve disputes, or maintain peace in the community. They can include initiatives by people in the community .” Baseline Evaluation | Shejeh Salam Activity 52 Female Youth 0.039 0.043 ** Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Older 0.000 0.487 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.167 0.033 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.111 0.214 Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 0.211 0.540 Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.115 0.003 ** Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.278 0.010 ** Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.347 Dinka 0.350 0.952 Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 0.615 0.051 * Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG FACTOR Estimate P-value Significance Pairwise analysis Male Older 0.500 0.251 Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.302 0.089 * Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 0.250 0.159 Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.395 0.435 Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.500 0.357 Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.444 0.475 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG Female Youth 0.351 0.292 Female, youth, Luo, in WBEG compared to all other Luo in WBEG Female Older 0.375 0.785 Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.251 Dinka 0.225 0.751 Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.375 0.379 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Indicator CPI (a) Indicator: Percent of community members who answer “yes” to the following question: “As far as you know, over the past year have there been any new peacebuilding systems or projects in your community? These can include people or groups who try and prevent conflict, resolve disputes, or maintain peace in the community. They can include initiatives by people in the community .” Baseline Evaluation | Shejeh Salam Activity 53 Male Older 0.158 0.201 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.333 0.062 * Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 0.059 0.243 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.386 0.009 ** Nuer in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.364 0.821 Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.214 0.023 ** Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.382 0.947 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.483 0.194 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.033 0.000 ** Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.091 0.540 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.048 0.755 Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.030 0.926 Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.000 0.117 Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.417 Dinka 0.302 0.104 Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.241 0.558 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.176 0.038 ** Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.382 0.250 Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.370 0.104 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.525 0.091 * Nuer in Upper Nile compared to all other ethnic groups in Upper Nile FACTOR Estimate P-value Significance Pairwise analysis Male Youth 0.625 0.243 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.540 0.808 Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator CPI (a) Indicator: Percent of community members who answer “yes” to the following question: “As far as you know, over the past year have there been any new peacebuilding systems or projects in your community? These can include people or groups who try and prevent conflict, resolve disputes, or maintain peace in the community. They can include initiatives by people in the community .” Baseline Evaluation | Shejeh Salam Activity 54 Female Youth 0.443 0.171 Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.565 0.621 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator CPI (b) Indicator: To measure indicator CPI (b), we take the average confidence levels – (1) confidence in their ability to prevent conflict from escalating into violence, and (2) confidence in their ability to resolve conflict after it has become violent. Baseline Evaluation | Shejeh Salam Activity 55 FACTOR Estimate P-value Significance Pairwise analysis Overall 0.76 Southern Unity 0.772 0.892 Southern Unity compared to all other regional clusters Eastern Eq. 0.699 0.599 Eastern Eq. compared to all other regional clusters WBEG 0.854 0.034 ** WBEG compared to all other regional clusters Jonglei 0.756 0.870 Jonglei compared to all other regional clusters Upper Nile 0.739 0.679 Upper Nile compared to all other regional clusters Dinka 0.920 0.000 ** Dinka compared to all other ethnic groups Nuer 0.709 0.226 Nuer compared to all other ethnic groups Luo 0.808 0.368 Luo compared to all other ethnic groups Murle 0.621 0.005 ** Murle compared to all other ethnic groups Buya 0.446 0.000 ** Buya compared to all other ethnic groups Toposa 0.951 0.000 ** Toposa compared to all other ethnic groups Other 0.878 0.030 ** Other compared to all other ethnic groups Male_Youth 0.792 0.071 * Male youth compared to all other demographic groups Male_Older 0.750 0.278 Male non-youth compared to all other demographic groups Female_Youth 0.765 0.906 Female youth compared to all other demographic groups Female_Older 0.753 0.638 Female non-youth compared to all other demographic groups Southern Unity 0.772 Dok Nuer 0.704 0.315 Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.716 0.555 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.740 0.196 Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator CPI (b) Indicator: To measure indicator CPI (b), we take the average confidence levels – (1) confidence in their ability to prevent conflict from escalating into violence, and (2) confidence in their ability to resolve conflict after it has become violent. Baseline Evaluation | Shejeh Salam Activity 56 Female Youth 0.656 0.091 * Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Older 0.662 0.524 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.627 0.014 ** Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.594 0.619 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 0.618 0.716 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.638 0.607 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.632 0.638 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 0.987 0.000 ** Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 1.000 0.321 Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 0.977 0.727 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 0.981 0.335 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 1.000 0.357 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.699 Buya 0.443 0.000 ** Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.528 0.123 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.453 0.837 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 0.383 0.057 * Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Indicator CPI (b) Indicator: To measure indicator CPI (b), we take the average confidence levels – (1) confidence in their ability to prevent conflict from escalating into violence, and (2) confidence in their ability to resolve conflict after it has become violent. Baseline Evaluation | Shejeh Salam Activity 57 Female Older 0.490 0.495 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.951 0.000 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.961 0.331 Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 0.980 0.053 * Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.947 0.851 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.917 0.369 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.854 Dinka 0.890 0.177 Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 0.942 0.359 Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 0.750 0.031 ** Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.943 0.007 ** Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 0.802 0.034 ** Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.808 0.074 * Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.773 0.432 Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.875 0.264 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG FACTOR Estimate P-value Significance Pairwise analysis Female Youth 0.811 0.899 Female, youth, Luo, in WBEG compared to all other Luo in WBEG Female Older 0.789 0.668 Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.756 Dinka 0.923 0.000 ** Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.906 0.815 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.987 0.072 * Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.907 0.595 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Indicator CPI (b) Indicator: To measure indicator CPI (b), we take the average confidence levels – (1) confidence in their ability to prevent conflict from escalating into violence, and (2) confidence in their ability to resolve conflict after it has become violent. Baseline Evaluation | Shejeh Salam Activity 58 Female Older 0.882 0.260 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.756 0.998 Nuer in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.818 0.238 Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.768 0.716 Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.772 0.527 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.707 0.118 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.621 0.001 ** Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.614 0.890 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.667 0.471 Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.576 0.074 * Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.644 0.533 Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.739 Dinka 0.934 0.000 ** Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.940 0.856 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.907 0.111 Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.968 0.034 ** Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.921 0.366 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.557 0.000 ** Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.531 0.546 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.583 0.547 Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.520 0.460 Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.573 0.730 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator CPI (c) Indicator: Respondents who strongly agree with the following statement: The use of violence is never justified Baseline Evaluation | Shejeh Salam Activity 59 FACTOR Estimate P-value Significance Pairwise analysis Overall 0.39 Southern Unity 0.660 0.001 ** Southern Unity compared to all other regional clusters Eastern Eq. 0.151 0.000 ** Eastern Eq. compared to all other regional clusters WBEG 0.428 0.552 WBEG compared to all other regional clusters Jonglei 0.237 0.016 ** Jonglei compared to all other regional clusters Upper Nile 0.465 0.179 Upper Nile compared to all other regional clusters Dinka 0.449 0.318 Dinka compared to all other ethnic groups Nuer 0.481 0.064 * Nuer compared to all other ethnic groups Luo 0.388 0.974 Luo compared to all other ethnic groups Murle 0.352 0.801 Murle compared to all other ethnic groups Buya 0.088 0.000 ** Buya compared to all other ethnic groups Toposa 0.217 0.076 * Toposa compared to all other ethnic groups Other 0.456 0.419 Other compared to all other ethnic groups Male_Youth 0.323 0.061 * Male youth compared to all other demographic groups Male_Older 0.343 0.066 * Male non-youth compared to all other demographic groups Female_Youth 0.434 0.061 * Female youth compared to all other demographic groups Female_Older 0.411 0.500 Female non-youth compared to all other demographic groups Southern Unity 0.660 Dok Nuer 0.469 0.031 ** Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.448 0.848 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.462 0.913 Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Youth 0.406 0.446 Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator CPI (c) Indicator: Respondents who strongly agree with the following statement: The use of violence is never justified Baseline Evaluation | Shejeh Salam Activity 60 Female Older 0.647 0.210 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.568 0.363 Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.500 0.745 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 0.647 0.044 ** Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.660 0.339 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.447 0.024 ** Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 0.958 0.000 ** Nyuong Nuer in Southern Unity compared Unity compared to other ethnic groups in Southern Unity Male Youth 1.000 0.063 * Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 1.000 0.073 * Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 0.969 0.638 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 0.912 0.457 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.151 Buya 0.088 0.305 Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.056 0.441 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.093 0.903 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 0.091 0.919 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Older 0.120 0.551 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.217 0.285 Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Indicator CPI (c) Indicator: Respondents who strongly agree with the following statement: The use of violence is never justified Baseline Evaluation | Shejeh Salam Activity 61 Male Youth 0.044 0.097 * Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 0.342 0.215 Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.148 0.165 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.417 0.214 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.428 Dinka 0.460 0.604 Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 0.538 0.475 Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 0.200 0.070 * Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.509 0.072 * Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 0.417 0.464 Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.388 0.503 Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.409 0.848 Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.500 0.251 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG FACTOR Estimate P-value Significance Pairwise analysis Female Youth 0.439 0.419 Female, youth, Luo, in WBEG compared to all other Luo in WBEG Female Older 0.219 0.010 ** Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.237 Dinka 0.239 0.976 Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.375 0.631 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.211 0.457 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.296 0.369 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 0.118 0.074 * Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.176 0.222 Nuer in Jonglei compared to all other ethnic groups in Jonglei Indicator CPI (c) Indicator: Respondents who strongly agree with the following statement: The use of violence is never justified Baseline Evaluation | Shejeh Salam Activity 62 Male Youth 0.273 0.206 Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.250 0.199 Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.147 0.234 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.138 0.313 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.352 0.349 Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.091 0.181 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.238 0.034 ** Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.364 0.208 Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.538 0.004 ** Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.465 Dinka 0.519 0.225 Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.379 0.130 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.431 0.228 Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.618 0.040 ** Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.574 0.401 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.426 0.345 Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.375 0.752 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.270 0.009 ** Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.525 0.119 Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.500 0.176 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator CPI (d) Indicator: Perceptions of safety in the respondent’s boma Baseline Evaluation | Shejeh Salam Activity 63 FACTOR Estimate P-value Significance Pairwise analysis Overall 0.55 Southern Unity 0.497 0.481 Southern Unity compared to all other regional clusters Eastern Eq. 0.367 0.270 Eastern Eq. compared to all other regional clusters WBEG 0.872 0.000 ** WBEG compared to all other regional clusters Jonglei 0.451 0.129 Jonglei compared to all other regional clusters Upper Nile 0.583 0.705 Upper Nile compared to all other regional clusters Dinka 0.707 0.031 ** Dinka compared to all other ethnic groups Nuer 0.497 0.371 Nuer compared to all other ethnic groups Luo 0.826 0.000 ** Luo compared to all other ethnic groups Murle 0.368 0.011 ** Murle compared to all other ethnic groups Buya 0.022 0.000 ** Buya compared to all other ethnic groups Toposa 0.714 0.009 ** Toposa compared to all other ethnic groups Other 0.884 0.000 ** Other compared to all other ethnic groups Male_Youth 0.580 0.348 Male youth compared to all other demographic groups Male_Older 0.530 0.131 Male non-youth compared to all other demographic groups Female_Youth 0.554 0.980 Female youth compared to all other demographic groups Female_Older 0.558 0.865 Female non-youth compared to all other demographic groups Southern Unity 0.497 Dok Nuer 0.465 0.664 Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.448 0.485 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.481 0.376 Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Youth 0.453 0.105 Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator CPI (d) Indicator: Perceptions of safety in the respondent’s boma Baseline Evaluation | Shejeh Salam Activity 64 Female Older 0.471 0.901 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.360 0.048 ** Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.469 0.084 * Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 0.353 0.800 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.340 0.139 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.342 0.584 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 0.665 0.015 ** Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.875 0.024 ** Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 0.545 0.216 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 0.669 0.807 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 0.647 0.583 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.367 Buya 0.017 0.000 ** Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.014 0.842 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.035 0.126 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 0.013 0.645 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Older 0.000 0.321 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.714 0.000 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Indicator CPI (d) Indicator: Perceptions of safety in the respondent’s boma Baseline Evaluation | Shejeh Salam Activity 65 Male Youth 0.644 0.270 Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 0.750 0.460 Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.738 0.407 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.722 0.793 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.872 Dinka 0.895 0.303 Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 0.962 0.093 * Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 1.000 0.000 ** Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.858 0.045 ** Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 0.896 0.973 Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.826 0.062 * Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.727 0.080 * Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.917 0.126 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG Female Youth 0.833 0.810 Female, youth, Luo, in WBEG compared to all other Luo in WBEG FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.828 0.949 Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.451 Dinka 0.289 0.000 ** Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.250 0.625 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.368 0.170 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.222 0.268 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 0.324 0.356 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.582 0.000 ** Nuer in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.477 0.069 * Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Indicator CPI (d) Indicator: Perceptions of safety in the respondent’s boma Baseline Evaluation | Shejeh Salam Activity 66 Male Older 0.643 0.286 Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.603 0.446 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.569 0.515 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.368 0.083 * Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.364 0.958 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.286 0.295 Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.439 0.036 ** Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.346 0.401 Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.583 Dinka 0.765 0.000 ** Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.776 0.833 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.735 0.450 Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.791 0.360 Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.759 0.802 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.421 0.000 ** Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.563 0.025 ** Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.437 0.667 Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.352 0.080 * Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.435 0.658 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator CPI (e) Indicator: In each county, the main conflicts – and the primary actors / ethnic groups – were identified. To measure interactions or relationships with previously strained or conflicting groups, we asked respondents about the frequency of marriage, trade, and shared water sources between groups. Baseline Evaluation | Shejeh Salam Activity 67 FACTOR Estimate P-value Significance Pairwise analysis Overall 0.75 Southern Unity 0.886 0.000 ** Southern Unity compared to all other regional clusters Eastern Eq. 0.622 0.000 ** Eastern Eq. compared to all other regional clusters WBEG 0.793 0.093 * WBEG compared to all other regional clusters Jonglei 0.660 0.003 ** Jonglei compared to all other regional clusters Upper Nile 0.789 0.149 Upper Nile compared to all other regional clusters Dinka 0.792 0.081 * Dinka compared to all other ethnic groups Nuer 0.799 0.034 ** Nuer compared to all other ethnic groups Luo 0.772 0.538 Luo compared to all other ethnic groups Murle 0.618 0.000 ** Murle compared to all other ethnic groups Buya 0.587 0.000 ** Buya compared to all other ethnic groups Toposa 0.657 0.001 ** Toposa compared to all other ethnic groups Other 0.781 0.265 Other compared to all other ethnic groups Male_Youth 0.738 0.224 Male youth compared to all other demographic groups Male_Older 0.747 0.592 Male non-youth compared to all other demographic groups Female_Youth 0.752 0.884 Female youth compared to all other demographic groups Female_Older 0.763 0.227 Female non-youth compared to all other demographic groups Southern Unity 0.886 Dok Nuer 0.796 0.000 ** Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.778 0.386 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.776 0.090 * Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator CPI (e) Indicator: In each county, the main conflicts – and the primary actors / ethnic groups – were identified. To measure interactions or relationships with previously strained or conflicting groups, we asked respondents about the frequency of marriage, trade, and shared water sources between groups. Baseline Evaluation | Shejeh Salam Activity 68 Female Youth 0.823 0.117 Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Older 0.837 0.233 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.951 0.017 ** Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.986 0.002 ** Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 0.941 0.664 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.946 0.693 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.947 0.906 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 0.920 0.203 Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 1.000 0.003 ** Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 0.970 0.097 * Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 0.901 0.113 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 0.922 0.958 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.622 Buya 0.586 0.021 ** Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.605 0.293 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.566 0.199 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 0.596 0.346 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Indicator CPI (e) Indicator: In each county, the main conflicts – and the primary actors / ethnic groups – were identified. To measure interactions or relationships with previously strained or conflicting groups, we asked respondents about the frequency of marriage, trade, and shared water sources between groups. Baseline Evaluation | Shejeh Salam Activity 69 Female Older 0.564 0.325 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.657 0.032 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.622 0.087 * Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 0.694 0.004 ** Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.647 0.197 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.679 0.318 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.793 Dinka 0.830 0.025 ** Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 0.829 0.962 Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 0.878 0.020 ** Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.825 0.556 Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 0.819 0.553 Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.772 0.273 Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.762 0.723 Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.826 0.025 ** Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG FACTOR Estimate P-value Significance Pairwise analysis Female Youth 0.786 0.283 Female, youth, Luo, in WBEG compared to all other Luo in WBEG Female Older 0.720 0.012 ** Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.660 Dinka 0.745 0.001 ** Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.722 0.292 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.749 0.571 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.741 0.788 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Indicator CPI (e) Indicator: In each county, the main conflicts – and the primary actors / ethnic groups – were identified. To measure interactions or relationships with previously strained or conflicting groups, we asked respondents about the frequency of marriage, trade, and shared water sources between groups. Baseline Evaluation | Shejeh Salam Activity 70 Female Older 0.758 0.453 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.629 0.152 Nuer in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.672 0.232 Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.643 0.524 Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.619 0.561 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.618 0.486 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.618 0.154 Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.605 0.717 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.611 0.743 Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.625 0.559 Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.620 0.945 Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.789 Dinka 0.791 0.895 Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.778 0.459 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.780 0.018 ** Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.800 0.071 * Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.799 0.392 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.786 0.854 Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.868 0.107 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.798 0.530 Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.728 0.069 * Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.807 0.386 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator CPI (full index) Indicator: Index score of 5 CPI components (a-e). Baseline Evaluation | Shejeh Salam Activity 71 FACTOR Estimate P-value Significance Pairwise analysis Overall 0.82 Southern Unity 0.905 0.156 Southern Unity compared to all other regional clusters Eastern Eq. 0.608 0.025 ** Eastern Eq. compared to all other regional clusters WBEG 0.971 0.001 ** WBEG compared to all other regional clusters Jonglei 0.713 0.039 ** Jonglei compared to all other regional clusters Upper Nile 0.883 0.209 Upper Nile compared to all other regional clusters Dinka 0.925 0.032 ** Dinka compared to all other ethnic groups Nuer 0.842 0.574 Nuer compared to all other ethnic groups Luo 0.944 0.021 ** Luo compared to all other ethnic groups Murle 0.621 0.000 ** Murle compared to all other ethnic groups Buya 0.423 0.000 ** Buya compared to all other ethnic groups Toposa 0.795 0.678 Toposa compared to all other ethnic groups Other 0.956 0.012 ** Other compared to all other ethnic groups Male_Youth 0.802 0.361 Male youth compared to all other demographic groups Male_Older 0.790 0.038 ** Male non-youth compared to all other demographic groups Female_Youth 0.825 0.485 Female youth compared to all other demographic groups Female_Older 0.836 0.308 Female non-youth compared to all other demographic groups Southern Unity 0.905 Dok Nuer 0.776 0.028 ** Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.756 0.429 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.781 0.703 Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Youth 0.762 0.518 Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator CPI (full index) Indicator: Index score of 5 CPI components (a-e). Baseline Evaluation | Shejeh Salam Activity 72 Female Older 0.819 0.038 ** Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.854 0.425 Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.894 0.254 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 0.845 0.715 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.863 0.144 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.830 0.324 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 1.093 0.000 ** Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 1.168 0.060 * Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 1.114 0.700 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 1.084 0.343 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 1.086 0.334 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.608 Buya 0.420 0.000 ** Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.436 0.476 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.414 0.752 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 0.418 0.867 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Older 0.415 0.846 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.795 0.000 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Indicator CPI (full index) Indicator: Index score of 5 CPI components (a-e). Baseline Evaluation | Shejeh Salam Activity 73 Male Youth 0.717 0.041 ** Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 0.865 0.039 ** Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.773 0.241 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.859 0.278 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.971 Dinka 1.005 0.082 * Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 1.094 0.005 ** Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 1.015 0.800 Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 1.002 0.797 Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 0.958 0.039 ** Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.944 0.262 Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.922 0.578 Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 1.054 0.009 ** Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG Female Youth 0.960 0.369 Female, youth, Luo, in WBEG compared to all other Luo in WBEG FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.862 0.018 ** Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.713 Dinka 0.733 0.449 Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.758 0.300 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.751 0.371 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.739 0.798 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 0.691 0.112 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.743 0.101 Nuer in Jonglei compared to all other ethnic groups in Jonglei Indicator CPI (full index) Indicator: Index score of 5 CPI components (a-e). Baseline Evaluation | Shejeh Salam Activity 74 Male Youth 0.758 0.592 Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.749 0.820 Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.742 0.955 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.735 0.606 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.621 0.002 ** Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.535 0.228 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.583 0.041 ** Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.642 0.276 Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.667 0.111 Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.883 Dinka 0.957 0.049 ** Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.914 0.158 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.904 0.037 ** Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 1.007 0.014 ** Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.978 0.347 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.814 0.049 ** Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.902 0.206 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.792 0.495 Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.770 0.369 Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.855 0.390 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator 2 Indicator: Custom Indicator 2 measures the percent of local community members reporting increased trust towards those that they have had a strained or past conflict with in the past 18 months. In each county, the main conflicts – and the primary actors / ethnic groups – were identified. Respondents were asked to comment on the trends in violence and trust between these conflict actors. Custom Indicator 2 measures the share of respondents who say “the conflict has become less violence than it Baseline Evaluation | Shejeh Salam Activity 75 used to be” and that trust levels between the conflicting groups have increased. FACTOR Estimate P-value Significance Pairwise analysis Overall 0.29 Southern Unity 0.247 0.716 Southern Unity compared to all other regional clusters Eastern Eq. 0.005 0.000 ** Eastern Eq. compared to all other regional clusters WBEG 0.347 0.357 WBEG compared to all other regional clusters Jonglei 0.416 0.030 ** Jonglei compared to all other regional clusters Upper Nile 0.407 0.196 Upper Nile compared to all other regional clusters Dinka 0.631 0.000 ** Dinka compared to all other ethnic groups Nuer 0.221 0.270 Nuer compared to all other ethnic groups Luo 0.310 0.798 Luo compared to all other ethnic groups Murle 0.549 0.001 ** Murle compared to all other ethnic groups Buya 0.000 0.000 ** Buya compared to all other ethnic groups Toposa 0.011 0.000 ** Toposa compared to all other ethnic groups Other 0.288 0.986 Other compared to all other ethnic groups Male_Youth 0.263 0.470 Male youth compared to all other demographic groups Male_Older 0.323 0.212 Male non-youth compared to all other demographic groups Female_Youth 0.272 0.479 Female youth compared to all other demographic groups Female_Older 0.290 0.888 Female non-youth compared to all other demographic groups Southern Unity 0.247 Dok Nuer 0.685 0.000 ** Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.759 0.517 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.731 0.119 Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator 2 Indicator: Custom Indicator 2 measures the percent of local community members reporting increased trust towards those that they have had a strained or past conflict with in the past 18 months. In each county, the main conflicts – and the primary actors / ethnic groups – were identified. Respondents were asked to comment on the trends in violence and trust between these conflict actors. Custom Indicator 2 measures the share of respondents who say “the conflict has become less violence than it Baseline Evaluation | Shejeh Salam Activity 76 Female Youth 0.563 0.058 * Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Older 0.647 0.802 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.017 0.080 * Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.000 0.498 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 0.000 0.482 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity used to be” and that trust levels between the conflicting groups have increased. FACTOR Estimate P-value Significance Pairwise analysis Female Youth 0.021 0.528 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Older 0.026 0.511 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 0.000 0.052 * Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.000 Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 0.000 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 0.000 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 0.000 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.005 Buya 0.000 0.110 Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.000 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.000 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 0.000 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Indicator 2 Indicator: Custom Indicator 2 measures the percent of local community members reporting increased trust towards those that they have had a strained or past conflict with in the past 18 months. In each county, the main conflicts – and the primary actors / ethnic groups – were identified. Respondents were asked to comment on the trends in violence and trust between these conflict actors. Custom Indicator 2 measures the share of respondents who say “the conflict has become less violence than it Baseline Evaluation | Shejeh Salam Activity 77 Female Older 0.000 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.011 0.114 Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.022 0.573 Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 0.000 0.154 Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.000 0.167 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.028 0.598 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.347 Dinka 0.470 0.144 Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 0.385 0.501 Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 0.300 0.130 Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.491 0.652 Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG used to be” and that trust levels between the conflicting groups have increased. FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.542 0.394 Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.310 0.600 Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.091 0.025 ** Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.444 0.175 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG Female Youth 0.351 0.462 Female, youth, Luo, in WBEG compared to all other Luo in WBEG Female Older 0.313 0.972 Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.416 Dinka 0.521 0.141 Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.625 0.541 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Indicator 2 Indicator: Custom Indicator 2 measures the percent of local community members reporting increased trust towards those that they have had a strained or past conflict with in the past 18 months. In each county, the main conflicts – and the primary actors / ethnic groups – were identified. Respondents were asked to comment on the trends in violence and trust between these conflict actors. Custom Indicator 2 measures the share of respondents who say “the conflict has become less violence than it Baseline Evaluation | Shejeh Salam Activity 78 Male Older 0.579 0.434 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.519 0.980 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 0.412 0.377 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.324 0.029 ** Nuer in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.182 0.178 Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.286 0.709 Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.294 0.545 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.431 0.145 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.549 0.037 ** Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.545 0.983 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.524 0.809 Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.606 0.582 Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.500 0.582 Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.407 Dinka 0.767 0.000 ** Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.828 0.234 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.824 0.208 Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.691 0.006 ** Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.759 0.846 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.084 0.000 ** Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.063 0.771 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile used to be” and that trust levels between the conflicting groups have increased. FACTOR Estimate P-value Significance Pairwise analysis Male Older 0.048 0.220 Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator 2 Indicator: Custom Indicator 2 measures the percent of local community members reporting increased trust towards those that they have had a strained or past conflict with in the past 18 months. In each county, the main conflicts – and the primary actors / ethnic groups – were identified. Respondents were asked to comment on the trends in violence and trust between these conflict actors. Custom Indicator 2 measures the share of respondents who say “the conflict has become less violence than it Baseline Evaluation | Shejeh Salam Activity 79 Female Youth 0.148 0.044 ** Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.065 0.535 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator 3 Indicator: Respondents were asked whether they know anyone in the past year who has been affected by: cattle raids, revenge killings, rape, and child abductions. The indicator is calculated based on the mean of these four indicators – or as a representation of the number of times a respondent said “yes” they know someone who was affected by the violence type. The score is between 0 and 1, where a score of 0 indicates a respondent said “no” they do not know anyone who has been affected by that, Baseline Evaluation | Shejeh Salam Activity 80 to all four violence types. A score of 1 indicates a respondent said “yes” to all four violence types. FACTOR Estimate P-value Significance Pairwise analysis Overall 0.45 Southern Unity 0.508 0.297 Southern Unity compared to all other regional clusters Eastern Eq. 0.563 0.004 ** Eastern Eq. compared to all other regional clusters WBEG 0.178 0.000 ** WBEG compared to all other regional clusters Jonglei 0.575 0.015 ** Jonglei compared to all other regional clusters Upper Nile 0.423 0.684 Upper Nile compared to all other regional clusters Dinka 0.320 0.028 ** Dinka compared to all other ethnic groups Nuer 0.528 0.036 ** Nuer compared to all other ethnic groups Luo 0.215 0.000 ** Luo compared to all other ethnic groups Murle 0.536 0.371 Murle compared to all other ethnic groups Buya 0.607 0.000 ** Buya compared to all other ethnic groups Toposa 0.508 0.175 Toposa compared to all other ethnic groups Other 0.172 0.000 ** Other compared to all other ethnic groups Male_Youth 0.431 0.428 Male youth compared to all other demographic groups Male_Older 0.464 0.525 Male non-youth compared to all other demographic groups Female_Youth 0.457 0.612 Female youth compared to all other demographic groups Female_Older 0.435 0.518 Female non-youth compared to all other demographic groups Southern Unity 0.508 Dok Nuer 0.410 0.302 Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.457 0.258 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.389 0.084 * Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator 3 Indicator: Respondents were asked whether they know anyone in the past year who has been affected by: cattle raids, revenge killings, rape, and child abductions. The indicator is calculated based on the mean of these four indicators – or as a representation of the number of times a respondent said “yes” they know someone who was affected by the violence type. The score is between 0 and 1, where a score of 0 indicates a respondent said “no” they do not know anyone who has been affected by that, Baseline Evaluation | Shejeh Salam Activity 81 Female Youth 0.398 0.751 Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Older 0.412 0.964 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.604 0.111 Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.641 0.535 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 0.662 0.540 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity to all four violence types. A score of 1 indicates a respondent said “yes” to all four violence types. FACTOR Estimate P-value Significance Pairwise analysis Female Youth 0.596 0.525 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Older 0.572 0.429 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 0.519 0.868 Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.531 0.845 Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 0.386 0.295 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 0.604 0.089 * Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 0.397 0.062 * Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.563 Buya 0.610 0.083 * Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.521 0.003 ** Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.610 0.999 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 0.643 0.069 * Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Indicator 3 Indicator: Respondents were asked whether they know anyone in the past year who has been affected by: cattle raids, revenge killings, rape, and child abductions. The indicator is calculated based on the mean of these four indicators – or as a representation of the number of times a respondent said “yes” they know someone who was affected by the violence type. The score is between 0 and 1, where a score of 0 indicates a respondent said “no” they do not know anyone who has been affected by that, Baseline Evaluation | Shejeh Salam Activity 82 Female Older 0.640 0.443 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.508 0.035 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.511 0.965 Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 0.500 0.869 Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.553 0.088 * Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.438 0.342 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.178 Dinka 0.140 0.217 Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 0.135 0.929 Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 0.200 0.254 Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.146 0.617 Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG to all four violence types. A score of 1 indicates a respondent said “yes” to all four violence types. FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.104 0.204 Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.215 0.247 Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.352 0.155 Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.236 0.577 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG Female Youth 0.175 0.329 Female, youth, Luo, in WBEG compared to all other Luo in WBEG Female Older 0.180 0.385 Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.575 Dinka 0.835 0.000 ** Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.938 0.014 ** Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.855 0.702 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Indicator 3 Indicator: Respondents were asked whether they know anyone in the past year who has been affected by: cattle raids, revenge killings, rape, and child abductions. The indicator is calculated based on the mean of these four indicators – or as a representation of the number of times a respondent said “yes” they know someone who was affected by the violence type. The score is between 0 and 1, where a score of 0 indicates a respondent said “no” they do not know anyone who has been affected by that, Baseline Evaluation | Shejeh Salam Activity 83 Female Youth 0.778 0.085 * Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 0.853 0.499 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.464 0.008 ** Nuer in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.636 0.007 ** Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.500 0.354 Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.438 0.321 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.414 0.154 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.536 0.644 Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.386 0.065 * Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.452 0.213 Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.561 0.654 Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.635 0.083 * Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.423 Dinka 0.214 0.000 ** Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.216 0.984 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.304 0.216 Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.141 0.216 Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.204 0.802 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.619 0.000 ** Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.703 0.345 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile to all four violence types. A score of 1 indicates a respondent said “yes” to all four violence types. FACTOR Estimate P-value Significance Pairwise analysis Male Older 0.563 0.479 Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.652 0.491 Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator 3 Indicator: Respondents were asked whether they know anyone in the past year who has been affected by: cattle raids, revenge killings, rape, and child abductions. The indicator is calculated based on the mean of these four indicators – or as a representation of the number of times a respondent said “yes” they know someone who was affected by the violence type. The score is between 0 and 1, where a score of 0 indicates a respondent said “no” they do not know anyone who has been affected by that, Baseline Evaluation | Shejeh Salam Activity 84 Female Older 0.621 0.974 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator 9 Indicator:Respondents were asked whether they think that traditional authorities are effective in maintaining peace in the community. Baseline Evaluation | Shejeh Salam Activity 85 FACTOR Estimate P-value Significance Pairwise analysis Overall 0.78 Southern Unity 0.938 0.000 ** Southern Unity compared to all other regional clusters Eastern Eq. 0.769 0.900 Eastern Eq. compared to all other regional clusters WBEG 0.736 0.376 WBEG compared to all other regional clusters Jonglei 0.696 0.153 Jonglei compared to all other regional clusters Upper Nile 0.771 0.849 Upper Nile compared to all other regional clusters Dinka 0.920 0.001 ** Dinka compared to all other ethnic groups Nuer 0.756 0.534 Nuer compared to all other ethnic groups Luo 0.674 0.132 Luo compared to all other ethnic groups Murle 0.736 0.577 Murle compared to all other ethnic groups Buya 0.555 0.000 ** Buya compared to all other ethnic groups Toposa 0.983 0.000 ** Toposa compared to all other ethnic groups Other 0.720 0.438 Other compared to all other ethnic groups Male_Youth 0.787 0.891 Male youth compared to all other demographic groups Male_Older 0.810 0.220 Male non-youth compared to all other demographic groups Female_Youth 0.787 0.759 Female youth compared to all other demographic groups Female_Older 0.748 0.126 Female non-youth compared to all other demographic groups Southern Unity 0.938 Dok Nuer 0.946 0.834 Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.931 0.712 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.942 0.836 Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Youth 0.938 0.704 Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator 9 Indicator:Respondents were asked whether they think that traditional authorities are effective in maintaining peace in the community. Baseline Evaluation | Shejeh Salam Activity 86 Female Older 1.000 0.316 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.864 0.166 Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.875 0.824 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 0.824 0.417 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.915 0.394 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.816 0.287 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 1.000 0.055 * Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 1.000 Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 1.000 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 1.000 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 1.000 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.769 Buya 0.552 0.000 ** Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.444 0.147 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.581 0.664 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 0.597 0.298 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Older 0.520 0.727 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.983 0.000 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Indicator 9 Indicator:Respondents were asked whether they think that traditional authorities are effective in maintaining peace in the community. Baseline Evaluation | Shejeh Salam Activity 87 Male Youth 0.978 0.068 * Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 1.000 0.390 Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.984 0.843 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.972 0.437 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.736 Dinka 0.830 0.047 ** Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 0.846 0.869 Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 0.700 0.289 Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.925 0.008 ** Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 0.667 0.083 * Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.674 0.212 Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.591 0.244 Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.778 0.310 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG Female Youth 0.632 0.352 Female, youth, Luo, in WBEG compared to all other Luo in WBEG FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.750 0.198 Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.696 Dinka 0.972 0.000 ** Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 1.000 0.128 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.947 0.614 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.963 0.817 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 1.000 0.130 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.545 0.015 ** Nuer in Jonglei compared to all other ethnic groups in Jonglei Indicator 9 Indicator:Respondents were asked whether they think that traditional authorities are effective in maintaining peace in the community. Baseline Evaluation | Shejeh Salam Activity 88 Male Youth 0.727 0.074 * Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.714 0.035 ** Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.529 0.704 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.414 0.017 ** Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.736 0.633 Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.818 0.294 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.952 0.032 ** Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.667 0.184 Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.615 0.066 * Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.771 Dinka 0.947 0.000 ** Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 1.000 0.219 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.863 0.135 Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.982 0.133 Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.963 0.223 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.609 0.000 ** Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.688 0.535 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.746 0.095 * Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.459 0.012 ** Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.597 0.880 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator 19a Percent of community members with adequate knowledge of trauma – including both causes and symptoms or effects of trauma. Baseline Evaluation | Shejeh Salam Activity 89 FACTOR Estimate P-value Significance Pairwise analysis Overall 0.80 Southern Unity 0.995 0.000 ** Southern Unity compared to all other regional clusters Eastern Eq. 0.997 0.000 ** Eastern Eq. compared to all other regional clusters WBEG 0.964 0.001 ** WBEG compared to all other regional clusters Jonglei 0.598 0.010 ** Jonglei compared to all other regional clusters Upper Nile 0.492 0.000 ** Upper Nile compared to all other regional clusters Dinka 0.620 0.023 ** Dinka compared to all other ethnic groups Nuer 0.800 0.932 Nuer compared to all other ethnic groups Luo 0.922 0.025 ** Luo compared to all other ethnic groups Murle 0.549 0.001 ** Murle compared to all other ethnic groups Buya 0.995 0.000 ** Buya compared to all other ethnic groups Toposa 1.000 0.000 ** Toposa compared to all other ethnic groups Other 0.952 0.002 ** Other compared to all other ethnic groups Male_Youth 0.870 0.015 ** Male youth compared to all other demographic groups Male_Older 0.726 0.021 ** Male non-youth compared to all other demographic groups Female_Youth 0.817 0.471 Female youth compared to all other demographic groups Female_Older 0.811 0.753 Female non-youth compared to all other demographic groups Southern Unity 0.995 Dok Nuer 0.992 0.679 Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 1.000 0.397 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.981 0.404 Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Youth 1.000 0.372 Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator 19a Percent of community members with adequate knowledge of trauma – including both causes and symptoms or effects of trauma. Baseline Evaluation | Shejeh Salam Activity 90 Female Older 1.000 0.376 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.992 0.491 Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 1.000 0.235 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 1.000 0.269 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.979 0.244 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Female Older 1.000 0.271 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 1.000 0.102 Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 1.000 Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 1.000 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 1.000 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 1.000 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.997 Buya 0.994 0.000 ** Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 1.000 0.620 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 1.000 0.578 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 1.000 0.391 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Older 0.960 0.012 ** Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 1.000 0.000 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Indicator 19a Percent of community members with adequate knowledge of trauma – including both causes and symptoms or effects of trauma. Baseline Evaluation | Shejeh Salam Activity 91 Male Youth 1.000 Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 1.000 Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 1.000 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 1.000 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.964 Dinka 0.980 0.232 Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 1.000 0.165 Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 0.900 0.396 Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.981 0.937 Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 1.000 0.183 Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.922 0.032 ** Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 1.000 0.033 ** Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.889 0.722 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG Female Youth 0.912 0.769 Female, youth, Luo, in WBEG compared to all other Luo in WBEG FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.906 0.619 Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.598 Dinka 0.014 0.000 ** Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.000 0.350 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.000 0.340 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.000 0.317 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 0.059 0.264 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.903 0.000 ** Nuer in Jonglei compared to all other ethnic groups in Jonglei Indicator 19a Percent of community members with adequate knowledge of trauma – including both causes and symptoms or effects of trauma. Baseline Evaluation | Shejeh Salam Activity 92 Male Youth 0.864 0.607 Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.893 0.850 Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.868 0.306 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.966 0.029 ** Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.549 0.633 Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.636 0.561 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.571 0.808 Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.424 0.124 Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.654 0.149 Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.492 Dinka 0.651 0.040 ** Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.552 0.225 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.569 0.037 ** Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.745 0.035 ** Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.685 0.518 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.356 0.045 ** Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.438 0.599 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.222 0.008 ** Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.344 0.848 Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.484 0.041 ** Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator 19b Indicator: Of the community members with adequate knowledge of trauma, the percent who are able to relate trauma with aggressive behavior and conflict. Baseline Evaluation | Shejeh Salam Activity 93 FACTOR Estimate P-value Significance Pairwise analysis Overall 0.74 Southern Unity 0.984 0.000 ** Southern Unity compared to all other regional clusters Eastern Eq. 0.997 0.000 ** Eastern Eq. compared to all other regional clusters WBEG 0.878 0.025 ** WBEG compared to all other regional clusters Jonglei 0.377 0.000 ** Jonglei compared to all other regional clusters Upper Nile 0.462 0.000 ** Upper Nile compared to all other regional clusters Dinka 0.587 0.069 * Dinka compared to all other ethnic groups Nuer 0.682 0.322 Nuer compared to all other ethnic groups Luo 0.783 0.614 Luo compared to all other ethnic groups Murle 0.549 0.015 ** Murle compared to all other ethnic groups Buya 0.995 0.000 ** Buya compared to all other ethnic groups Toposa 1.000 0.000 ** Toposa compared to all other ethnic groups Other 0.928 0.001 ** Other compared to all other ethnic groups Male_Youth 0.823 0.004 ** Male youth compared to all other demographic groups Male_Older 0.691 0.185 Male non-youth compared to all other demographic groups Female_Youth 0.746 0.586 Female youth compared to all other demographic groups Female_Older 0.701 0.205 Female non-youth compared to all other demographic groups Southern Unity 0.984 Dok Nuer 0.977 0.408 Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.931 0.483 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.981 0.874 Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Youth 1.000 0.084 * Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator 19b Indicator: Of the community members with adequate knowledge of trauma, the percent who are able to relate trauma with aggressive behavior and conflict. Baseline Evaluation | Shejeh Salam Activity 94 Female Older 1.000 0.072 * Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.983 0.942 Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 1.000 0.235 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 1.000 0.269 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.957 0.244 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Female Older 1.000 0.271 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 0.992 0.341 Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 1.000 0.528 Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 1.000 0.528 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 0.985 0.543 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 1.000 0.542 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.997 Buya 0.994 0.000 ** Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 1.000 0.620 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 1.000 0.578 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 1.000 0.391 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Older 0.960 0.012 ** Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 1.000 0.000 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Indicator 19b Indicator: Of the community members with adequate knowledge of trauma, the percent who are able to relate trauma with aggressive behavior and conflict. Baseline Evaluation | Shejeh Salam Activity 95 Male Youth 1.000 Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 1.000 Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 1.000 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 1.000 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.878 Dinka 0.920 0.394 Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 1.000 0.177 Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 0.900 0.748 Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.906 0.649 Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 0.917 0.949 Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.783 0.099 * Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.909 0.055 * Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.778 0.959 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG Female Youth 0.772 0.903 Female, youth, Luo, in WBEG compared to all other Luo in WBEG FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.719 0.581 Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.377 Dinka 0.014 0.000 ** Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.000 0.350 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.000 0.340 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.000 0.317 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 0.059 0.264 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.460 0.164 Nuer in Jonglei compared to all other ethnic groups in Jonglei Indicator 19b Indicator: Of the community members with adequate knowledge of trauma, the percent who are able to relate trauma with aggressive behavior and conflict. Baseline Evaluation | Shejeh Salam Activity 96 Male Youth 0.545 0.540 Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.536 0.316 Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.456 0.936 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.397 0.450 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.549 0.020 ** Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.636 0.561 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.571 0.808 Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.424 0.124 Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.654 0.149 Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.462 Dinka 0.619 0.038 ** Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.552 0.426 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.529 0.053 * Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.727 0.021 ** Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.630 0.823 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.327 0.041 ** Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.375 0.734 Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.222 0.027 ** Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.295 0.591 Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.452 0.047 ** Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Indicator 19c Indicator: Of the community members with adequate knowledge of trauma, the percent reporting that there are some trauma related interventions in the community Baseline Evaluation | Shejeh Salam Activity 97 FACTOR Estimate P-value Significance Pairwise analysis Overall 0.21 Southern Unity 0.209 0.963 Southern Unity compared to all other regional clusters Eastern Eq. 0.264 0.592 Eastern Eq. compared to all other regional clusters WBEG 0.211 0.902 WBEG compared to all other regional clusters Jonglei 0.198 0.885 Jonglei compared to all other regional clusters Upper Nile 0.151 0.268 Upper Nile compared to all other regional clusters Dinka 0.179 0.598 Dinka compared to all other ethnic groups Nuer 0.209 0.923 Nuer compared to all other ethnic groups Luo 0.233 0.741 Luo compared to all other ethnic groups Murle 0.055 0.001 ** Murle compared to all other ethnic groups Buya 0.038 0.000 ** Buya compared to all other ethnic groups Toposa 0.494 0.000 ** Toposa compared to all other ethnic groups Other 0.168 0.397 Other compared to all other ethnic groups Male_Youth 0.207 0.966 Male youth compared to all other demographic groups Male_Older 0.193 0.473 Male non-youth compared to all other demographic groups Female_Youth 0.202 0.768 Female youth compared to all other demographic groups Female_Older 0.222 0.370 Female non-youth compared to all other demographic groups Southern Unity 0.209 Dok Nuer 0.254 0.748 Dok Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.207 0.508 Male, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Male Older 0.269 0.048 ** Male, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Female Youth 0.250 0.937 Female, youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Indicator 19c Indicator: Of the community members with adequate knowledge of trauma, the percent reporting that there are some trauma related interventions in the community Baseline Evaluation | Shejeh Salam Activity 98 Female Older 0.294 0.477 Female, non-youth, Dok Nuer, in Southern Unity compared to all other Dok Nuer in Southern Unity Haak Nuer 0.000 0.018 ** Haak Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.000 Male, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Male Older 0.000 Male, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Female Youth 0.000 Female, youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.000 Female, non-youth, Haak Nuer, in Southern Unity compared to all other Haak Nuer in Southern Unity Nyuong Nuer 0.356 0.118 Nyuong Nuer in Southern Unity compared to other ethnic groups in Southern Unity Male Youth 0.250 0.309 Male, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Male Older 0.273 0.436 Male, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Youth 0.385 0.416 Female, youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Female Older 0.353 0.969 Female, non-youth, Nyuong Nuer, in Southern Unity compared to all other Nyuong Nuer in Southern Unity Eastern Equatoria 0.264 Buya 0.033 0.000 ** Buya in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Male Youth 0.028 0.842 Male, youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Male Older 0.047 0.578 Male, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Youth 0.039 0.709 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Female Older 0.000 0.321 Female, non-youth, Buya, in Eastern Equatoria compared to all other Buya in Eastern Equatoria Toposa 0.494 0.000 ** Toposa in Eastern Equatoria compared to all other ethnic groups in Eastern Equatoria Indicator 19c Indicator: Of the community members with adequate knowledge of trauma, the percent reporting that there are some trauma related interventions in the community Baseline Evaluation | Shejeh Salam Activity 99 Male Youth 0.333 0.039 ** Male, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Male Older 0.579 0.364 Male, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Youth 0.508 0.242 Female, youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria Female Older 0.583 0.275 Female, non-youth, Toposa, in Eastern Equatoria compared to all other Toposa in Eastern Equatoria WBEG 0.211 Dinka 0.190 0.640 Dinka in WBEG compared to all other ethnic groups in WBEG Male Youth 0.231 0.679 Male, youth, Dinka, in WBEG compared to all other Dinka in WBEG Male Older 0.200 0.911 Male, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Youth 0.189 0.971 Female, youth, Dinka, in WBEG compared to all other Dinka in WBEG Female Older 0.167 0.659 Female, non-youth, Dinka, in WBEG compared to all other Dinka in WBEG Luo 0.233 0.687 Luo in WBEG compared to all other ethnic groups in WBEG Male Youth 0.318 0.401 Male, youth, Luo, in WBEG compared to all other Luo in WBEG Male Older 0.167 0.263 Male, non-youth, Luo, in WBEG compared to all other Luo in WBEG Female Youth 0.211 0.548 Female, youth, Luo, in WBEG compared to all other Luo in WBEG FACTOR Estimate P-value Significance Pairwise analysis Female Older 0.250 0.777 Female, non-youth, Luo, in WBEG compared to all other Luo in WBEG Jonglei 0.198 Dinka 0.000 0.000 ** Dinka in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.000 Male, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Male Older 0.000 Male, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.000 Female, youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Female Older 0.000 Female, non-youth, Dinka, in Jonglei compared to all other Dinka in Jonglei Nuer 0.364 0.000 ** Nuer in Jonglei compared to all other ethnic groups in Jonglei Indicator 19c Indicator: Of the community members with adequate knowledge of trauma, the percent reporting that there are some trauma related interventions in the community Baseline Evaluation | Shejeh Salam Activity 100 Male Youth 0.364 1.000 Male, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Male Older 0.214 0.046 ** Male, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.353 0.864 Female, youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Female Older 0.448 0.205 Female, non-youth, Nuer, in Jonglei compared to all other Dinka in Jonglei Murle 0.055 0.011 ** Murle in Jonglei compared to all other ethnic groups in Jonglei Male Youth 0.091 0.613 Male, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Male Older 0.095 0.298 Male, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Youth 0.000 0.092 * Female, youth, Murle, in Jonglei compared to all other Dinka in Jonglei Female Older 0.077 0.549 Female, non-youth, Murle, in Jonglei compared to all other Dinka in Jonglei Upper Nile 0.151 Dinka 0.233 0.100 Dinka in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.172 0.414 Male, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Male Older 0.176 0.139 Male, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Youth 0.309 0.155 Female, youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Female Older 0.241 0.884 Female, non-youth, Dinka, in Upper Nile compared to all other Dinka in Upper Nile Nuer 0.079 0.116 Nuer in Upper Nile compared to all other ethnic groups in Upper Nile Male Youth 0.375 0.038 ** Male, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Male Older 0.079 0.996 Male, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Youth 0.016 0.042 ** Female, youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Female Older 0.065 0.798 Female, non-youth, Nuer, in Upper Nile compared to all other Nuer in Upper Nile Baseline Evaluation | Shejeh Salam Activity 101 ANNEX I1: SAMPLED BOMAS The following bomas were included in the final sample. Regional Cluster County Boma Eastern Equatoria Budi Ngarich Eastern Equatoria Kapoeta North Choroi Eastern Equatoria Kapoeta North Logerain Eastern Equatoria Kapoeta North Lolepan Eastern Equatoria Kapoeta North Morunyang Jonglei Akobo Gakdong Jonglei Akobo Kier Jonglei Akobo Mer Jonglei Akobo Nukta Jonglei Akobo Okau Jonglei Akobo Old Akobo Jonglei Akobo Padida Jonglei Akobo Rine Jonglei Akobo Weijiokni Jonglei Duk Amiel Jonglei Duk Dhuony Jonglei Duk Padiet Jonglei Duk Tindiir Jonglei Pibor Kongor Jonglei Pibor Kotile Jonglei Pibor Lawol Jonglei Pibor Lekuangole Jonglei Pibor Manydhing Jonglei Pibor Murlil Jonglei Pibor Wunkok Jonglei Uror Dhuony (Wikol Payam) Jonglei Uror Jokrial B (from Motot) Jonglei Uror Karam Jonglei Uror Pathai Jonglei Uror Riangchukuel Jonglei Uror Tangnyang (Motot Payam) Southern Unity Leer Dok Southern Unity Leer Norkong Southern Unity Leer Padeah Southern Unity Leer Payak Baseline Evaluation | Shejeh Salam Activity 102 Southern Unity Mayendit Daplual Southern Unity Mayendit Leah/Dhornyiel Southern Unity Mayendit Rubkuay Southern Unity Panyijar Kol Southern Unity Panyijar Nyal Southern Unity Panyijar Tiap Regional Cluster County Boma Upper Nile Baliet Akoka Centre Upper Nile Baliet Duut Upper Nile Baliet Pagak Upper Nile Baliet Rom Centre Upper Nile Baliet Wekjur Upper Nile Baliet Wunthow (Wunthow) Upper Nile Ulang Bul Buong Upper Nile Ulang Kuich Upper Nile Ulang Riang Upper Nile Ulang Rup-Buot Upper Nile Ulang Toalare Upper Nile Ulang Tor Kiel Upper Nile Ulang Waaga Upper Nile Ulang Ying Western Bahr-el Ghazal Jur River Achot Western Bahr-el Ghazal Jur River Alel Chok Western Bahr-el Ghazal Jur River Alur Western Bahr-el Ghazal Jur River Gumel Western Bahr-el Ghazal Jur River Kangi Western Bahr-el Ghazal Jur River Kayango Western Bahr-el Ghazal Jur River Kuajena Western Bahr-el Ghazal Jur River Kuom Western Bahr-el Ghazal Jur River Mabior Abiem Western Bahr-el Ghazal Jur River Marial Bai Western Bahr-el Ghazal Jur River Nyinakok Western Bahr-el Ghazal Jur River Rakbak Western Bahr-el Ghazal Jur River Rocrocdong Western Bahr-el Ghazal Jur River Tharkueng Western Bahr-el Ghazal Wau Agok Western Bahr-el Ghazal Wau Aweil Jedid Western Bahr-el Ghazal Wau Hai Bafara Western Bahr-el Ghazal Wau Hai Fallata Baseline Evaluation | Shejeh Salam Activity 103 Western Bahr-el Ghazal Wau Hai Kalvario Western Bahr-el Ghazal Wau Hai Krash A Western Bahr-el Ghazal Wau Hai Krash B Western Bahr-el Ghazal Wau Jebel A Western Bahr-el Ghazal Wau Jebel Kheir Western Bahr-el Ghazal Wau Kosti Western Bahr-el Ghazal Wau Muwzifin Western Bahr-el Ghazal Wau Nazareth Western Bahr-el Ghazal Wau New Site Western Bahr-el Ghazal Wau Salaam ANNEX III: COUNTY CONFLICTS The table below specifies the conflict participants were asked to respond to for questions related to Custom Indicator 2 and Community Peace Index component (e). Conflict Label Southern Unity Mayendit Nuer Haak Leer Nuer Dok Panyijiar Nuer Nyuong Eastern Equatoria Kapoeta North Toposa and Boya Budi Ddinga, Boya, Toposa Western Bahr-el Ghazal Wau Luo, Fertit, Dinka Jur River Luo and Dinka Jonglei Akobo Lou Nuer, Jikany Nuer, Anywak, Murle Duk Dinka and Nuer Pibor Murle, Jie Baseline Evaluation | Shejeh Salam Activity 104 Uror Lou Nuer Upper Nile Baliet Dinka Abiliang Ulang Nuer Jikany ANNEX IV: OVERVIEW OF LOCAL CONFLICT DYNAMICS38 Southern Unity Mayendit 2008 Census population: 53,783 2020 Population projection: 66,015 Ethnic groups and languages: Haak Nuer (Beek, Jalok, Kuey) Displacement Figures: 7,576 IDPs and 13,983 returnees (2020) Conflict dynamics in Mayendit include cattle raiding and intercommunal clashes. The Conflict Sensitivity Resource Facility (CSRF) in South Sudan notes that in northern Mayendit, ‘a group of armed cattle camp youth known as Tahariir often spar with youth from southern Mayendit and Leer County’. These Mayendit youth groups from also engage in cattle raids and low-level conflict with other youths from Panyijar County and Rumbek North, especially cattle-keepers from Pakam Payam. Though there have been multiple rounds of negotiations and community consultations with neighboring communities to resolve these conflicts, conflict has continued. At the national level, when the December 2013 violence erupted, Mayendit County and most of southern Unity, were predominantly SPLA-IO strongholds and the SPLA-IO still maintains a presence in the region, and northern Mayendit are impacted by IO-affiliated armed youth and community militias. Collectively, these national and local dynamics continue to contribute to instability in the region, which is exacerbated by IDP returns to Mayendit South from Panyijar County. Leer 2008 Census population: 53,022 2020 Population projection: 59,120 Ethnic groups and languages: Nuer (Adok) Baseline Evaluation | Shejeh Salam Activity 105 Displacement Figures: 12,376 IDPs and 11,916 returnees (2020) Pre-December 2013, local conflicts were the primary conflict driver in Leer, including cattle raiding among Leer, Mayendit, and Panyijar county youths. 2012 community consultations suggests that the drivers of these conflicts included food insecurity, water, and grazing competition, and a contested post-2011 independence returnee process. The CSRF notes that other factors included competition over the taxation of the main road to the county’s oil fields. Leer was particularly badly impacted by the 2013 conflict and is symbolically key to the IO as it is Riek Machar’s birthplace. The region is now SSPDF-controlled though the IO maintains a presence in neighbouring parts of the county, including the strategically vital Port of Adok. Panyijar 2008 Census population: 50,723 38 For additional details please refer to the Conflict Sensitivity Resource Facility (CSRF) Country Profiles, from which much of this text has been adapted from: https://www.csrf-southsudan.org/county-profiles/ 2020 Population projection: 109,541 Ethnic groups and languages: Nuer (Nyuong) Displacement Figures: 10,684 IDPs and 6,174 returnees (2020) Both violence and cooperation has frequently taken place between Dinka and Nuer communities in the region and cattle raiding and conflicts related to cattle grazing are one of the main forms of insecurity. Panyijar has largely been an IO stronghold since December 2013. The CSRF notes that the SPLA-IO is also suspected of recruiting child soldiers in the region. Baseline Evaluation | Shejeh Salam Activity 106 Eastern Equatoria Kapoeta North 2008 Census population: 103,084 2020 Population projection*: 153,352 Ethnic groups and languages: Toposa Displacement Figures: 5,948 IDPs and 1,469 returnees (2020) Conflict dynamics in Mayendit include cattle raiding and intercommunal clashes among groups from Warrap, The Conflict Sensitivity Resource Facility in South Sudan notes that in northern Mayendit, ‘a group of armed cattle camp youth known as Tahariir often spar with youth from southern Mayendit and Leer County’. These Mayendit youth groups from also engage in cattle raids and low-level conflict with other youths from Panyijar County and Rumbek North (especially cattle-keepers from Pakam Payam). Though there have been multiple rounds of negotiations and community consultations with neighboring communities to resolve these conflicts, conflict has continued. At the national level, when the December 2013 violence erupted, Mayendit County and most of southern Unity, were predominantly SPLA-IO strongholds and the SPLA-IO still maintains a presence in the region, and northern Mayendit are impacted by IO-affiliated armed youth and community militias. Collectively, these national and local dynamics continue to contribute to instability in the region, which is exacerbated by IDP returns to Mayendit South from Panyijar County. Budi 2008 Census population: 99,234 2020 Population projection: 99,242 Ethnic groups and languages: Didinga and Buya Displacement Figures: 3,314 IDPs and 3,412 returnees (2020) From at least the 1980s, there have been multiple sources of violence in what is now Budi County. Chukudum served as the SPLA headquarters in 1992, which also led to an influx of IDPs to the region. Divisions within the SPLA, known as the ‘Chukudum crisis’ contributed to violence that was resolved in a 2002 local peace agreement. Additionally, there are tensions between Budi County residents and Logiri and Toposa cattle rustlers. More recently, following the 2013 civil war, the SPLA clashed with civilians and there were skirmishes between the SPLA and alleged deserters. There have also been intercommunal attacks from the Lords Resistance Army (LRA) and local communities between the 2005 signing of the CPA and independence. Baseline Evaluation | Shejeh Salam Activity 107 Western Bahr el Ghazal Wau 2008 Census population: 151,320 2020 Population projection: 314,949 Major ethnic groups and languages: Balanda Viri, Balanda Bor, Luo/‘Jur Chol’, Dinka (Marial Bai) Displacement Figures: 46,555 IDPS and 149,010 returnees (2019) According to the Small Arms Survey, there have been tensions between the SPLA and local ethnic minorities, which are often referred to as the ‘Fertit’, in the region since the 1980s. The term Fertit initially referred to migration from Darfur, but now applies more broadly to non-Dinka groups in Western Bahr el-Ghazal. Some groups for whom the term is applied self-refer to themselves as Fertit while others, such as the Balanda apply their pre-existing ethnic or tribal names. Tensions between different local communities and migrant communities and the SPLA contributed to selfdefence forces in the region to ward off raids in Raja and Wau Counties. There have also been protests over the shift of the capital from Wau County to Wau Town prior to the December 2013 civil war. There have been waves of violence in the region since December 2013. In 2014 alleged SPLA defections of Nuer soldiers contributed to instability, the following year cattle keepers in Wau County contributed to conflicts as did clashes between opposition forces and the government. Local armed groups, sometimes operating as the ‘Fertit Lions’ or the ‘Western Bahr el Ghazal Lions’ emerged in response and have had shifting alliances with the SPLAIO. Government forces and the opposition clashed in 2017, which contributed to the displacement of civilians and fighting has persisted in 2018, too. The demographics of Wau town has subsequently been transformed by these waves of conflict, members of the Fertit and Luo are suspected to have been particularly impacted. Conflict dynamics in Wau are also shaped by displacement to the region from neighboring Jur River and Raja counties. The region has been relatively peaceful since early 2019, as the R-ARCSS has improved the security situation, though clashes between armed cattle keepers and farmers still contribute to displacement. Jur River 2008 Census population: 127,771 2020 Population projection: 273,118 Major ethnic groups and languages: Balanda Bor, Luo/’Jur Chol’. Displacement Figures: 16,838 IDPS and 34,777 returnees (2019) Baseline Evaluation | Shejeh Salam Activity 108 Conflict in Jur River County is predominantly driven by cattle raiding and pastoralist violence over land and water access and crop damage caused by livestock. These clashes have contributed to mass displacement to Wau and Tambura counties. A July 2019 tri-state conference sought to ameliorate tensions, but with limited success. Jonglei Akobo 2008 Census population: 136,210 2020 Population projection: 225,439 Ethnic groups and languages: Lou Nuer and Anyuak Displacement Figures: 16,948 IDPs and 17,304 returnees (2020) Contemporary violence in Akobo has roots in December 2013. The region was under SPLA-IO control through most of 2013-2016. In addition to the SPLA-IO, the region also home to the White Army (Jech Mabor), which is an informal network of Nuer community defence groups that is partly organised around community protection demands. On the whole, owing to the presence of prophets, the White Army is independent of major armed groups, though it has aligned with them. Communal clashes have also taken place in Akobo County. For instance, the CSRF notes that Murle communities have engaged in ‘large-scale violence, cattle raiding, and the abduction of children, all of which have been significant drivers of conflict’. UNMISS has attempted to implement local peace agreements in the region, with limited success as clashes have endured. The Luo Nuer and Anyuak have also been engaged in conflicts, in part because of the ways in which the latter have allegedly displaced the former and unresolved land disputes. Waves of returns have been taking place since 2019, which is a potential conflict driver. Duk 2008 Census population: 65,588 2020 Population projection: 189,958 Ethnic groups and languages: Dinka Hol (Angaach, Duoor, Nyieel, and Pathel), Dinka Nyarweng Displacement Figures: 22,191 IDPs and 37,187 returnees (2019) Baseline Evaluation | Shejeh Salam Activity 109 Duk County has primarily been under opposition control since December 2013, though some parts of eastern Duk County have been government controlled. Duk residents were displaced when the government reclaimed control of Bor town. Duk and Twic East counties are Dinka-majority and have maintained peaceful relations with Nuer neighboring communities and the International Crisis Group noted in 2014 that Dinka-Nuer relations were strongest in Duk County than in any other part of the country in part because of local peacebuilding initiatives. Nuer and Dinka fighting has largely sidestepped the region and instead been focused in Bor South County, though the White Army has contributed to some violence in Duk and there have been communal clashes between communities in Duk County and neighbouring counties. This includes clashes between the Murle from Pibor and Luo Nuer and Dinka communities in Duk. Duk County chiefs held a community peace dialogue in 2017, which was UNDP-supported and an an ‘open security forum’ was held in 2019 among regional stakeholders with UNMISS-support. Cattle raids and insecurity have continued despite these initiatives, which the CSRF argues are part of broader political contestation and efforts to acquire administrative control. Pibor 2008 Census population: 148,475 2020 Population projection: 204,879 Ethnic groups and languages: Murle, Kapchipo and Jie Displacement Figures: 30,166 IDPs and 11,153 returnees (2020) Pibor County conflict dynamics vary from other parts of the country and from other parts of Jonglei. In 2010 David Yau Yau established an armed opposition group to contest subnational leadership the region, under the broad claim that regional government authorities neglected the Murle. In 2011 Yau Yau signed a peace agreement with the Government of South Sudan, but he defected in 2012 and created the SSDM-Cobra Faction. Fighting between the SSDM-Cobra Faction and the Government of South Sudan contributed to mass displacement. According to the Conflict Sensitivity Resource Facility this timeframe also contributed to the ‘perceived instrumentalization of ethnicity by outside groups’, which has led to tensions within Murle communities. Pibor is also home to the Greater Pibor Administrative Area, which was in part forged through peace agreements between the Government of South Sudan and the SSDM-Cobra Faction as the latter was reluctant to fight a conflict in two areas. The move is attributed with ameliorating conflict in the region and preventing it from becoming a major national conflict-site, though there has still been large scale displacement in Pibor Town. In addition to conflicts tied to national dynamics, intracommunity conflicts have also been a component of Pibor’s local conflict dynamics. The CSRF notes that conflicts have been in part driven by cattle raids, child abductions, and revenge killing, which have become increasingly ensnared in ‘armed inter-communal and antigovernment conflicts and in new networks of criminality’. This includes conflicts between Murle age-sets, which has galvanised cooperation between the Luo Nuer and Bor Dinka against what the latter two ethnic groups ostensibly perceive as a rising Murle threat. Uror Baseline Evaluation | Shejeh Salam Activity 110 2008 Census population: 178,519 2020 Population projection: 182,012 Ethnic groups and languages: Lou Nuer Displacement Figures: 13,993 IDPs and 8,588 returnees (2020) The Lou Nuer dominates the population of Urur County and the region was primarily under SPLA-IO control following the December 2013 civil war. Community-driven efforts to prevent the region from becoming ensnared in the national conflict have been somewhat successful; for instance, in 2015 authorities from across Jonglei, including from Uror, collaborated to ensure the region maintained diplomatic ties with neighbouring counties. These initiatives included cross-border trading activities and grassroots peacebuilding efforts. The primary drivers of insecurity in Uror include communal clashes, notably between the Lou Nuer and the Murle from neighboring Pibor County, have limited the regions access to post-CPA infrastructure improvements and have led to the abduction of women and children, violence against civilians, cattle theft, and property destruction. The CSRF notes that revenge is a common driver of these acts of violence as well as grievances against the government following disarmament campaigns. Upper Nile Baliet 2008 Census population: 48,010 2020 Population projection: 54,058 Ethnic groups and languages: Dinka (Padang, Ngok Lual Yak, which is made up of eleven sub-clans, including: Ajuba, Awieer (Awiɛɛr), Adong (Adɔ̈ŋ), Achaak (Acaak), Abii, Baliët, Balak, Duut, Ding (Dïng), Ngaar (Ŋäär) and Dhiaak), Nuer and Shilluk Displacement Figures: 5,747 IDPs and 13,852 returnees (2020) Violence in Beliet County pre-dates the national civil war that emerged in December 2013. Conflicts have frequently been over land access, including grazing rights for livestock, and border disputes along the southern county boundary and on the White Nile between the Padang Dinka and Shilluk communities who reside on the Fashoda County boundary. Work by Joshua Craze notes that communities interpret the uptick in violence to rising tensions over state boundaries and national political interests that have become increasingly localised. Baliet County was also disrupted by the national conflict that emerged in December 2013. Government aligned local militias have emerged, initially to protect the Paloich oil fields and area communities. Though these militias Baseline Evaluation | Shejeh Salam Activity 111 have at times aligned with the SPLA they are not part of a formal SPLA command. Baliet County has also been esnared between SPLA and SPLA-IO conflicts and consequently received an influx of IDPs. The 2015 announcement of 10 to 28 states worsened tensions among Padang Dinka and Shilluk communities. Though South Sudan has returned to 10-states, there are lingering resentments, including that the Shilluk are the victims of a longstanding administrative and political campaign to forcibly remove them from the White Nile’s east banks. Ulang 2008 Census population: 85,044 2020 Population projection: 115,640 Ethnic groups and languages: Jikany Nuer (Cie-Lang: Nguandong, Khorchol, Lonythior, Walok Chany) Displacement Figures: 17,619 IDPs and 36,404 returnees (2020) Much of the violence in Ulang County is localised, including cattle-raiding and revenge killing between the Nyirol Lou Nuer and the Ulang County Nuer communities. Conflicts are often over grazing rights, which include both land and water. These disputes are exacerbated by the prevalence of small arms, disputed boundaries, displacement, and food insecurity. During the December 2013 violence, most of the Upper Nile Nuer population sided with the SPLA-IO, including Nuers in Ulang County the wider Upper Nile region subsequently became of the main conflict zones in the country, where disputes between the SPLA and SPLA-IO violently took place. Though the SSPDF (formerly the SPLA) now controls Nasir town, its control over the area is contested and relations between the SSPDF and the local population have been periodically violent. The 2015 government shift from 10 to 28 states contributed to a pattern whereby ethnic fiefdoms were instituted in the region, particularly among Nuer communities who were allotted their own Balkanized territories, which was reversed by the 2020 return to 10states. Though the R-ARCSS has calmed some tensions, intercommunal violence persists in the region. ANNEX V: RANKING OF NEWS SOURCE RELIABILITY Ranking of the sources based on the % of respondents reporting that it is the most trusted source % with radio Region access Relatives Tradition al Leaders Religious Leaders Radio Friends Political Leaders Social Media Southern Unity 1 3 2 5 4 6 6 13% Mayendit 1 4 2 5 3 6 7 10% Leer 2 1 5 4 3 7 6 18% Panyijar 2 4 1 5 3 6 7 12% Baseline Evaluation | Shejeh Salam Activity 112 Eastern Equatoria 1 2 5 4 3 6 7 7% Kapoeta North 2 1 5 4 3 6 7 12% Budi 1 3 6 4 2 5 7 1% Western Bahr el Ghazal 2 4 2 1 5 7 6 48% Wau 3 5 2 1 4 7 6 66% Jur River 2 4 3 1 5 7 6 31% Jonglei 1 2 3 6 4 5 7 11% Akobo 2 1 3 4 5 7 6 7% Duk 3 2 1 4 5 6 7 24% Pibor 1 2 4 6 5 3 7 80% Uror 1 2 3 6 4 5 7 40% Upper Nile 1 2 3 4 5 6 7 24% Baliet 1 3 2 4 5 6 7 30% Ulang 1 3 2 4 5 6 7 79% Key: Source where the % of respondents reporting that it is the most trusted source is ranked…. Key: Access to radio Highest 1 Over 70% Higher 2 60%- less than 70% High 3 50%- less than 60% Middle 4 40%- less than 50% Low 5 30%- less than 40% Lower 6 20%- less than 30% Lowest 7 Less than 20% Baseline Evaluation | Shejeh Salam Activity 113 U.S. Agency for International Development www.usaid.gov